Publications

Prompting fairness: Learning prompts for debiasing large language models

Authors: Camelia Lemnaru, Cristian Andrei Rad

Large language models are prone to internalize social biases due to the characteristics of the data used for their self-supervised training scheme. Considering their recent emergence and wide availability to the general public, it is mandatory to identify and alleviate these biases to avoid perpetuating stereotypes towards underrepresented groups. We present a novel prompt-tuning method for reducing biases in encoder models such as BERT or RoBERTa. Unlike other methods, we only train a small set of additional reusable token embeddings that can be concatenated to any input sequence to reduce bias in the outputs. We particularize this method to gender bias by providing a set of templates used for training the prompts. Evaluations on two benchmarks show that our method is on par with the state of the art while having a limited impact on language modeling ability

A Hybrid Machine Learning–Genetic Algorithm for Optimizing Surface-Mount Technology Planning

Authors: Adrian Petru Groza

Projects: HRIA - Romanian HUB for AI

We tackle the problem of improving the Surface- Mount Technology (SMT) process planning in an automotive manufacturing setting. Current simulations show low accu- racy across production lines as the existing approach relies on predefined setups rather than adapting to product-specific configurations. We propose a hybrid framework that couples machine learning with a genetic algorithm to generate product- specific plans. Our solution involves three tasks: (i) assigning boards to lines, (ii) allocating components to Pick-and-Place (PnP) machines, and (iii) balancing workloads across machines. Our hybrid pipeline embeds supervised learning in a genetic optimizer. A multi-class classifier selects feasible PnP head con- figurations per Bill of Materials (BOM) part number (precision = 0.73). A genetic algorithm assigns components to compatible feeder tables/machines, while a regression model estimates table cycle times (R² = 0.88). The fitness jointly optimizes Components Placed per Hour (CPH) and Line Balancing (LB) under process constraints. Different mutation methods are explored, revealing that mutation based on balancing the workload by leveling the number of placements on the tables with minimum and maximum cycle time results in an LB of 0.83, with a CPH of 0.37 and an average delta cycle time of -3.27% across 105-part numbers

ConvU-NExT: An Asymmetrical Encoder–Decoder for Denoising Low Dose CT

Authors: Adrian Petru Groza

Projects: HRIA - Romanian HUB for AI

Low-dose computed tomography (LDCT) is a medical imaging modality designed to minimize ionizing radiation exposure while maintaining the ability to produce detailed cross-sectional images. It is particularly valuable in scenarios requiring repeated imaging, such as cancer screening, follow-up examinations or pediatric diagnostics, where reducing radiation dose is critical to patientsafety. For example, to reduce noise by half, fourtimesthe radiation dose isrequired in the slice. The goal isto achieve postprocessed LDCT images with comparable quality to those obtained from standard-dose CT imaging. We start with a brief overview of the CT procedures and their limitations. Then we introduce a novel denoising method based on an asymmetric integration of the ConvNeXt backbone with the U-Net architecture. This novel approach obtained 2–3 times less noise than the original LDCT, having a 10%–20% increase in performance compared to U-Net implementation, checked against three metrics MSE, SSIMLoss and combinations of both. The results suggest that: (i) augmenting the images with specific noise, obtained from water phantom CT scan test, while training yieldssuperiorresults compared to generic noise augmentations; (ii) a larger kernelsize better extracts features and (iii) a smaller kernel size was mandatory for feature reconstruction

Review Before Trust: Source-Grounded Integrity Gates for AI-Assisted Personal Health Records

Authors: Adrian Petru Groza

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

Large language models can convert medical documents into structured data, but plausible output may still be unsupported by the source. Persisting such output in a longitudinal health record, a record that accumulates patient information over time, therefore creates an integrity risk: unverified data may influence later summaries, trends, or preventive-care computations. We introduce an evidence-gated trust-promotion model that keeps generated data provisional until a deterministic monitor verifies it against the source document. The monitor admits a candidate for a specified downstream use only when the source contains a unique supporting quotation, the relevant fields occur within the same laboratory row, and the required provenance is preserved. The generator cannot approve its own output, missing or ambiguous evidence causes refusal, and refused candidates remain available for human review rather than being silently discarded. We implement the model in Medical DataCloud, a personal health-record application, and evaluate it through automated tests and a replay of saved extraction outputs. All 22 conformance and mutation tests pass. The replay covers nine historical laboratory PDF reports containing 102 manually labelled rows. The reports produce 97 numeric candidates: schema validation accepts all 97, an earlier packet-level evidence check accepts 94, and the hardened quotation- and row-level policy admits 72 while retaining 25 for review. The study evaluates system integrity rather than clinical correctness or clinical safety. The results demonstrate the technical feasibility of an enforceable boundary that prevents generated claims from authorizing their own reuse in a longitudinal health record.

Memory-First Fact-Checking: A Knowledge-Graph-Grounded Multi-Agent System for Misinformation Detection

Authors: Alexandru Lecu, Adrian Petru Groza

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

This paper introduces a hybrid fact-checking framework that integrates Knowledge Graph-based semantic memory with adversarial multi-agent reasoning for explainable misinformation detection. The proposed system follows a memory-first, web-fallback architecture, in which input claims are initially evaluated against a dual-index Knowledge Graph through SentenceBERT-based semantic retrieval and Natural Language Inference. When the evidence retrieved from the graph is insufficient to support a reliable decision, the framework collects information from trusted web sources and assesses it using an adversarial tribunal composed of support, contradiction, and judging agents. A graphaware confidence mechanism combines semantic similarity, NLI confidence, and structural graph evidence to determine whether internal knowledge is sufficient, thereby reducing unnecessary web retrieval. Following verification, validated information is transformed into structured triples and incorporated into the Knowledge Graph, supporting the incremental expansion of the system’s semantic memory. Experimental evaluation on a curated COVID-19 misinformation benchmark demonstrates that the proposed framework achieves an accuracy of 97.4% and a macroaveraged F1-score of 92.6% on resolved claims, outperforming a Llama 3.3 70B baseline, which obtains an accuracy of 87.7% and a macro-averaged F1-score of 86.3%.

An Explainable Coherence Score for Detecting Temporal Inconsistencies in Political News

Authors: Adrian Petru Groza

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

Temporal inconsistencies, such as mandates attributed outside their real interval, events presented as past before they occurred, or inverted causal sequences, are a form of political disinformation that evades style-based fake news detectors: a well-written article with a single wrong date carries no lexical signal of falsehood. This paper introduces the Temporal Coherence Score (TCS), a continuous, intrinsically interpretable metric that quantifies the temporal coherence of a news article, computed by a four-stage pipeline: extraction of temporal facts, construction of a temporal knowledge graph, hierarchical verification against internal consistency rules and external reference sources, and score aggregation with automatically generated explanations. Verification combines eight internal checkers derived from Allen’s interval algebra with a five-level external hierarchy ranging from a locally stored reference knowledge base of 1,256 curated political facts to live Wikidata SPARQL queries. On a benchmark of 100 political news articles with injected temporal errors, the system reaches a precision of 0.909 at the selected operating threshold, with a single residual false positive, a profile deliberately tuned for human-in-the-loop fact-checking assistance, where false alarms are costlier than missed detections. Unlike lexical baselines that output only a binary label, every flagged article is accompanied by the inconsistency type, the entities involved, and the reference source that contradicts the claim.

Prompts and Prayers:the Rise of GPTheology

Authors: Ioana Cheres, Adrian Petru Groza

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

Increasingly artificial intelligence (AI) has been cast in “god-like” roles (to name a few: film industry – Matrix, The Creator, Mission Impossible, Foundation, Dune etc.; literature – Children of Time, Permutation City, Neuromancer, I Have no Mouth and I Must Scream, Alphaville etc.). This trend has accelerated with the advent of sophisticated Large Language Models such as ChatGPT. For this phenomenon, where AI is perceived as divine, we use the term GPTheology, where ChatGPT and other AI models are treated as potential oracles of a semi-divine nature. This paper explores the emergence of GPTheology as a form of techno-religion, examining how narratives around AI echo traditional religious constructs. We draw on community narratives from online forums – Reddit – and recent projects – AIpowered Mazu Statue in Malaysia (Lu, 2025); “ShamAIn” Project in Korea (He-rim, 2025); AI Jesus in a Swiss Church (Kennedy, 2024). These examples show striking similarities to technological notions of the Singularity and the development of Artificial General Intelligence (AGI). Additionally, we analyse how daily interactions with AI are acquiring ritualistic associations and how AI-centric ideologies clash with or are integrated into established religions. This study uses a dataset of Reddit posts discussing AI to identify recurring themes of salvation, prophecy, and demonization surrounding AI. Our findings suggest that new belief systems are developing around AI, and this carries both philosophical and sociotechnical implications. Our paper critically analyses the benefits and dangers, as well as the social, political and ethical challenges of this development. This transdisciplinary inquiry highlights how AI and religion are increasingly intertwined, prompting necessary questions about humanity’s relationship with its creations and the future of belief.

Agent-Assisted Composition of Adaptive Business Workflows: A Semantic-Driven Modular Architecture

Authors: Adrian Petru Groza

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

.

Towards glass-box triage: an interpretable multi-agent clinical decision support system for emergency department

Authors: Roxana Ramona Szomiu, Adrian Petru Groza

Projects: Adobe Learning Funds - Agentic AI

[Preprint] Towards glass-box triage: an interpretable multi-agent clinical decision support system for emergency department

Ontology-guided multi-agent reasoning for explainable differential diagnosis

Authors: Roxana Ramona Szomiu, Adrian Petru Groza

Projects: Adobe Learning Funds - Agentic AI

[Preprint]Ontology-guided multi-agent reasoning for explainable differential diagnosis

AlloyGraph: Data and Evaluation Results for Multi-Agent AI Superalloy Property Prediction

Authors: Alexandru Lecu, Adrian Petru Groza

Training data (77 alloys from the Nickel Institute handbook), evaluation data (88 alloys from manufacturer datasheets), prediction results for six model configurations, chatbot evaluation benchmarks (250 MCQ questions, 100 RAGAS questions, 12 expert-graded questions), inverse design results (20 target specifications), and OWL ontology for the AlloyGraph platform. Associated repository: https://github.com/AlexLecu/AlloyGraph

Reconciling Manufacturer Claims with Measured Degradation in Commercial Lithium-Ion Cells: A Provenance-Aware Knowledge Graph with Coverage-Gated Abstention

Authors: Alexandru Lecu, Adrian Petru Groza

Manufacturer datasheets state battery cycle life under conditions that rarely match how cells are used, while public cycling datasets measure degradation under conditions datasheets do not cover. We present a knowledge-graph (KG) system that represents claims, measurements, and independent tests of commercial lithium-ion cells with full provenance, detects claim-versus-measured and claim-versus-claim discrepancies conditioned on the comparability of test conditions, and supports cycle-life prediction with coverage-gated abstention. On the 124-cell Severson dataset under leave-one-policy-group-out cross-validation, graph-derived neighbor features do not significantly improve point prediction over a strong early-cycle baseline (RMSE 135 vs. 141 cycles), but graph coverage provides a statistically significant abstention signal (Spearman ρ=−0.25 with prediction error, p=0.006) that reduces retained RMSE by roughly 40% at 60% retention, where random abstention does not. Deployed zero-shot on a second cycling study of the same commercial cell, the gate abstained on all 77 cells; the counterfactual confirms every refusal (approximately 83% error had it answered), an error an ungated baseline commits silently. On a third study with commensurable features, the gate’s first partial acceptance (17 of 45 cells) is itself diagnostic: coverage acts partly as a lifetime proxy out of distribution, and five labeled cells halve retained error while leaving that proxy in place—adaptation repairs the predictor, not the selection criterion. A 70B open-weight LLM extracts datasheet claims at F1=0.70 with non-deterministic output even at temperature 0; a deterministic validator with three-run consensus raises this to F1=0.78 with zero unsourced values; on a held-out datasheet, precision and the zero-unsourced-value property transfer while recall falls to 0.34, localizing the extractor’s boundary at table-structured content; row-level table grounding, implemented in response, raises held-out recall to 0.63 with zero hallucinations at a measured precision cost. Reconciling claims across document variants shows that roughly one in three cross-document specification comparisons (14 of 43, three commercial cells) yields a conflict or condition mismatch, twelve involving third-party documents and two internal to a single manufacturer’s own documents. A hand-labeled, condition-annotated gold standard of 103 claims (62 development, 41 held-out; inter-annotator κ=0.74 on property naming) and a staged, human-gated literature-monitoring pipeline are released with the code.

Understanding the strengths and weaknesses of SSL models for audio deepfake model attribution

Authors: Adriana Stan

Audio deepfake model attribution aims to mitigate the misuse of synthetic speech by identifying the source model responsible for generating a given audio sample, enabling accountability and informing vendors. The task is challenging, but self-supervised learning (SSL)-derived acoustic features have demonstrated state-of-the-art attribution capabilities, yet the underlying factors driving their success and the limits of their discriminative power remain unclear. In this paper, we systematically investigate how SSL-derived features capture architectural signatures in audio deepfakes. By controlling multiple dimensions of the audio generation process we reveal how subtle perturbations in model checkpoints, text prompts, vocoders, or speaker identity influence attribution. Our results provide new insights into the robustness, biases, and limitations of SSL-based deepfake attribution, highlighting both its strengths and vulnerabilities in realistic scenarios.

AlloyGraph: A Knowledge-Graph-Grounded Multi-Agent System for Nickel-Based Superalloy Design

Authors: Alexandru Lecu

AMD Ontology — Data and Evaluation Results for Agentic Ontology Engineering using LLMs and DL-Learner

Authors: Alexandru Lecu

Clinical trial abstracts used for ontology construction, UMLS alignment evaluation results, DL-Learner class expression learning experiments (17 learning problems, CELOE/OCEL algorithms), competency question evaluation (143 AMD patient questions), and OWL 2 DL ontology for Age-Related Macular Degeneration.Associated repository: https://github.com/CalinaBorzan/AMD-ontology

Matrix effect resolution in artificial intelligence-driven bacteria detection

Authors: Anca Nicoleta Mărginean

Accurate detection of bacteria in complex biological, food, and environmental matrices is essential for infection control, antimicrobial resistance (AMR) surveillance, and public health decision-making. However, matrix effects arising from heterogeneous sample compositions remain a major barrier to the reliable deployment of rapid novel detection technologies. These effects can distort analytical signals, reduce reproducibility, and complicate data interpretation, particularly outside controlled laboratory settings. Recent advances in artificial intelligence (AI) offer new opportunities to overcome matrix-induced variability by enabling data-driven signal correction, pattern recognition, and adaptive learning across diverse sample types. In this review, we synthesized emerging strategies from the last 5 years for matrix-effect resolution in AI-assisted bacterial detection systems, including electrochemical, optical, and spectroscopic platforms. We discuss how AI models can integrate biological variability, environmental context, and sensor heterogeneity to improve robustness, analyte discrimination, and even quantitative accuracy, with emphasis on high-risk pathogens, such as those in the ESKAPE group. The workflow for AI-assisted bacterial detection is critically analyzed, beginning with pretreatment of complex samples, processing of raw data to extract the most relevant features, and then continuing with the implementation of the algorithm.

Contrastive Learning in Lesion Detection for Mammography Screening Programs Explained with XAI

Authors: Anca Nicoleta Mărginean

From ML Predictions to Informed Diagnostic Assistance Using the Toulmin Model of Argumentation

Authors: Anca Nicoleta Mărginean

To provide a structured and interpretable assessment, we decompose the image-based diagnosis into components following the Toulmin model of argumentation. This model consists of a claim, grounds, warrant, qualifier, rebuttal, and backing. Consider a claim generated by a machine learning (ML) model for retinal diagnosis. Rather than accepting this claim at face value, one could either apply explainable AI (XAI) methods or adopt an argumentation-based approach. In our framework, a model specialized in biomarker extraction from images provides the grounds. The warrant-linking the grounds to the claim - is analyzed by an agent equipped with medical knowledge; in our architecture, this role is fulfilled by a MedGemma agent. The qualifier is determined based on the overall quantitative evaluation of both the warrant and grounds models. Finally, a rebuttal is constructed using image similarity measures computed with MedSigLip. All these components are presented to the human expert, enabling a more informed and critical assessment of the ML-generated diagnosis.

Energy Forecasting under Missing Data: Comparative Evaluation of Augmented Representations and Decoder-only Time-Series Imputation

Authors: Tudor Cioara, Mircea Gabriel Antonesi, Anghel Ionuț

Projects: HRIA - Romanian HUB for AI

Data-related issues, including missing values and irregular measurements, challenge the accuracy of short-term energy forecasting in smart grids. In data-scarce scenarios, two approaches are commonly considered, but their strengths and weaknesses are not fully mapped. Embedding-based models learn joint representations from heterogeneous data, compensating for the lack of time-series measurements via additional contextual or external sources, whereas imputation pipelines restore temporal continuity but may smooth variability or produce implausible values. To address these limitations, we propose a unified forecasting framework for energy systems that integrates a shared Temporal Fusion Transformer prediction with a controlled degradation protocol to simulate realistic missing-data patterns. This enables a fair and systematic comparison between two pipelines: a representation-augmented learning and decoder-only time series imputation. The former integrates TS2Vec temporal embeddings and BERT-based static contextual representations to provide a richer forecasting space without explicit reconstruction of missing values. The latter uses a Chronos-2 model to reconstruct missing time-series segments, followed by physics-based correction to enforce physically plausible outputs. We evaluate both pipelines under a controlled data degradation protocol to map the trade-offs between representation learning and data continuity restoration through imputation. We use real-world non-residential building electricity consumption and wind generation datasets. The imputation-based pipeline achieves a mean sMAPE of 10.14% and MAE of 8.43 kWh across 100 buildings, compared to 12.11% and 10.89 kWh for the representation-based approach (p<0.01, p<0.01, p<0.01). On the wind generation dataset, imputation also improves predictive accuracy (R²=0.870 vs. R²=0.794). However, representation-based models remain competitive in scenarios with irregular, spike-dominated, or event-driven consumption patterns where imputation provides limited additional benefits.

P2P Energy Trading Coordination in Interconnected Microgrid Systems

Authors: Tudor Cioara, Anghel Ionuț

Performance Evaluation of LLMs in Automated RDF Knowledge Graph Generation

Authors: Tudor Cioara, Anghel Ionuț

Cloud systems generate large, heterogeneous log data containing critical infrastructure, application, and security information. Transforming these logs into RDF triples enables their integration into knowledge graphs, improving interpretability, root-cause analysis, and cross-service reasoning beyond what raw logs allow. Large Language Models (LLMs) offer a promising approach to automate RDF knowledge graph generation; however, their effectiveness on complex cloud logs remains largely unexplored. In this paper, we evaluate multiple LLM architectures and prompting strategies for automated RDF extraction using a controlled framework with two pipelines for systematically processing semi-structured log data. The extraction pipeline integrates multiple LLMs to identify relevant entities and relationships, automatically generating subject-predicate-object triples. These outputs are evaluated using a dedicated validation pipeline with both syntactic and semantic metrics to assess accuracy, completeness, and quality. Due to the lack of public ground-truth datasets, we created a reference Log-to-KG dataset from OpenStack logs using manual annotation and ontology-driven methods, enabling objective baseline. Our analysis shows that Few-Shot learning is the most effective strategy, with Llama achieving a 99.35% F1 score and 100% valid RDF output while Qwen, NuExtract, and Gemma also perform well under Few-Shot prompting, with Chain-of-Thought approaches maintaining similar accuracy. One-Shot prompting offers a lighter but effective alternative, while Zero-Shot and advanced strategies such as Tree-of-Thought, Self-Critique, and Generate-Multiple perform substantially worse. These results highlight the importance of contextual examples and prompt design for accurate RDF extraction and reveal model-specific limitations across LLM architectures.

A systematic review of generative AI usage for IT project management

Authors: Tudor Cioara, Anghel Ionuț

This paper aims to synthesize current knowledge on generative AI in IT project management using the PRISMA methodology to provide researchers with a comprehensive perspective on techniques, applications, adoption trends, limitations, and integration across project management tools and process groups. The analysis reveals a clear dominance of OpenAI's GPT in the included studies but relying primarily on prompt engineering, suggesting that research in this area remains at an exploratory stage. Finally, it identifies and discusses three promising research directions for AI-enabled project management, including process group-specific AI agents, project role-based AI agents, and hybrid collaborative networks that enable human-guided orchestration.

Performance evaluation of large language models for automated knowledge graph generation

Authors: Tudor Cioara, Anghel Ionuț

Cloud systems generate large and heterogeneous raw data logs containing information related to infrastructure, application, and security. Transforming these logs into RDF triples enables their integration into knowledge graphs, improving interpretability, root-cause analysis, and cross-service reasoning. Large Language Models (LLMs) offer a promising approach to automate RDF knowledge graph generation. However, their effectiveness on complex cloud logs remains largely unexplored. In this paper, we evaluate multiple LLM architectures and prompting strategies for automated RDF extraction using a controlled framework with two pipelines for systematically processing semi-structured log data. The extraction pipeline integrates multiple LLMs to identify relevant entities and relationships, automatically generating subject-predicate-object triples. These outputs are evaluated using a dedicated validation pipeline with both syntactic and semantic metrics to assess accuracy, completeness, and quality. Due to the lack of public ground-truth datasets, we created a reference Log-to-KG dataset from OpenStack logs using manual annotation and ontology-driven methods, enabling objective baseline. Few-Shot learning achieved the best overall performance in our evaluation, with LLaMA achieving a 99.35% F1 score and 100% valid RDF output while Qwen, NuExtract, and Gemma also perform well under Few-Shot prompting, with Chain-of-Thought approaches maintaining similar accuracy. One-Shot prompting offers a lighter but effective alternative, while Zero-Shot and strategies such as Single-Prompt Tree-of-Thought, and Generate-Multiple-Then-Vote showed consistently lower performance. The top-performing LLMs, LLaMA and Qwen, with One-Shot and Few-Shot prompting were evaluated on a different dataset, to assess transferability, performance stability and efficiency in terms of latency, token generation, and GPU memory usage. Our work highlights the importance of contextual examples and prompt design for accurate RDF extraction and reveal model-specific limitations across LLM architectures offering practical guidelines and a foundation for future knowledge graphs extraction from unstructured data research.

Memetic-based Coordination of Distributed Storage Units Flexibility for Congestion Management

Authors: Tudor Cioara, Anghel Ionuț, Viorica Chifu

Older people with mild cognitive impairment engaged by social robot-based intervention: Benefits shown in multicountry long term randomized controlled trial (Preprint)

Authors: Tudor Cioara, Anghel Ionuț

BACKGROUND Loneliness and social isolation are some of the several risk factors that contribute to dementia. Technology‑based interventions using socially assistive robots (SARs) and mobile apps may help not only to maintain cognitive functioning but also to support social connectedness and psychosocial wellbeing in older adults with mild cognitive impairment (MCI); however, evidence from multicountry trials remains limited. OBJECTIVE The proof‑of‑concept (PoC) evaluated the benefits of the engAGE platform, a hybrid intervention combining a social robot, mobile app, and wearable activity tracker, designed to support both social connectedness and cognitive functioning in older people with MCI across three European countries. METHODS Older adults with MCI were recruited in Italy, Switzerland, and Norway. The 6‑month intervention combined weekly robot‑guided group sessions with daily tablet use and continuous activity tracker wear at home. Outcomes were assessed at baseline and post-test and included subjective memory complaints as primary outcome (MAC‑Q), and global cognition (MoCA), loneliness (UCLA), quality of life (QoL‑AD and EQ‑5D‑5L VAS), and mental wellbeing (WEMWBS) as secondary outcomes. Also, usability and acceptance were assessed through SUS and UTAUT after 3 and 6 months of intervention. Intra-group and inter‑group differences in change were explored for any dimension to determine the effects of the intervention. RESULTS Of 50 enrolled participants, 44 (36 assigned to the experimental group - EG; and 8 to the control group - CG) completed the final assessment and were included in the analyses. The subjective memory complaints reduced significantly from 26.41 (±2.23) to 25.22 (±3.19) in the EG, whereas remained unchanged in the CG (26.75 ±0.71). MoCA scores remained stable overall (EG: 23.51 ±2.16 to 23.44 ±3.21), with no significant differences between groups. Psychosocial outcomes showed a mixed pattern: in the Italian EG, loneliness decreased significantly (UCLA: 45.18 ± 9.61 to 37.94 ± 6.56), whereas in Switzerland significantly worsened (p=.021). Also, the self‑rated health (EQ‑5D‑5L-VAS) improved significantly in the EG (p=.013), but no significant differences were detected between groups. Overall QoL‑AD and WEMWBS scores remained broadly stable. SUS scores in the EG improved significantly (p=.020) from 58.81 (±18.17) to 68.54 (±18.54), reaching the commonly accepted usability threshold. CONCLUSIONS The engAGE hybrid intervention, combining robot‑guided group activities with home‑based tablet use and activity monitoring, was delivered across three different socio‑healthcare contexts and showed preliminary benefits in subjective memory, selected psychosocial measures, and usability. CLINICALTRIAL ClinicalTrials.gov NCT06302686; https://clinicaltrials.gov/study/NCT06302686 INTERNATIONAL REGISTERED REPORT RR2-10.2196/67601

Generative AI for IT Project Management: A Systematic Review and Future Research Agenda

Authors: Tudor Cioara, Anghel Ionuț

Nowadays, the literature on Generative AI (GenAI) in Information Technology (IT) project management is fragmented, focusing mainly on isolated tools, specific process groups, or practitioners’ perspectives, without offering a comprehensive synthesis. Therefore, there is a lack of systematic reviews to guide researchers in effectively and responsibly leveraging GenAI, including emerging innovations such as AI agents. This paper aims to synthesize current knowledge on GenAI in IT project management, combining a PRISMA-compliant systematic review of the peer-reviewed literature, a complementary analysis of commercial and open-source platforms, and a forward-looking research agenda featuring our vision on agentic AI architectures for IT project management. For the systematic review based on academic sources we have used the Web of Science (WoS) database in our study. Studies were eligible if published between 2021 and 2026 in English, as journal articles or conference proceedings, across major publishers (IEEE, Springer, Elsevier, MDPI, ACM, and others), and indexed under computer science, engineering, or AI categories in WoS. For industry-driven analysis, sources included vendor documentation, official product pages, and publicly accessible repository specifications, selected for relevance through manual search. The review reveals that while academic research remains largely focused on prompt-based applications of foundation models such as GPT, commercial and open-source platforms have progressed toward embedding GenAI as an operational capability within project workflows. Therefore, we consider that agentic architecture represents a promising future direction for enabling autonomous task execution, collaborative decision-making, and human–AI orchestration and integration across the project lifecycle.

Asynchronous federated learning with partial weights aggregation for energy consumption forecasting

Authors: Anghel Ionuț, Tudor Cioara

Accurate energy forecasting is essential for grid stability, demand-side management, and efficient renewable integration. However, energy consumption data collected from smart meters may expose sensitive user information, thus raising privacy concerns. Federated Learning (FL) offers a privacy-preserving mechanism for collaborative model training without sharing raw data. However, conventional synchronous FL suffers from training delays caused by heterogeneous client availability and computational capabilities, while frequent exchange of model parameters can lead to communication overheads. To address these challenges, this paper proposes an asynchronous federated learning framework for energy forecasting that enables continuous global model updating without waiting for all clients to complete local training. We introduce a federated asynchronous adaptive aggregation mechanism, where client-specific learning rates are dynamically adjusted based on both update staleness and model performance contribution. A partial aggregation strategy is defined for a Long Short-Term Memory (LSTM) forecasting model that splits the local models' layers, allowing clients to exchange only a subset of the weights with the server. The proposed solution is evaluated using real-world energy consumption data from multiple consumers. Experimental results demonstrate that the proposed asynchronous adaptive strategy outperforms the classic FedAvg approach and maintains prediction accuracy relative to personalised FedAvg, while reducing communication costs. Additionally, the proposed method outperforms the classic FedAsync algorithm across all client groups, with statistically significant improvements in most cases.

Optimizing Appliances Usage for Apartment Buildings Participation in Demand Response Programs

Authors: Anghel Ionuț, Tudor Cioara

This paper presents a bio-inspired optimisation framework that enables apartment buildings to participate more effectively in demand response programmes by coordinating the use of household appliances while preserving residents' comfort. Developed within the context of the DEDALUS project, the proposed approach addresses the complexity of managing multiple apartments with heterogeneous appliances, demonstrating how collective flexibility can support grid stability without requiring the participation of every household. The proposed methodology formulates appliance scheduling as a bi-level optimisation problem, combining a Genetic Algorithm with the Harris Hawks Optimisation (HHO) algorithm to identify both the minimum number of apartments required for demand response participation and the optimal operating schedules for their appliances. The framework incorporates user preferences, appliance flexibility and comfort constraints to generate recommendations that closely match the target consumption profile requested by the distribution system operator while minimising disruptions to residents' daily routines. The approach is validated using data collected from a residential apartment building, demonstrating its ability to closely approximate target energy consumption profiles while involving only a subset of apartments. The experimental results highlight the algorithm's convergence, scalability and effectiveness in balancing energy flexibility with occupant comfort, providing a practical decision-support tool for integrating multi-apartment residential buildings into future smart grid and demand response ecosystems.

Distributed Joint Embedding Predictive Architecture for Energy Forecasting

Authors: Anghel Ionuț, Tudor Cioara

SAREF-based Ontology for Distributed AI Workflows across the Edge-Fog-Cloud Continuum

Authors: Vasile Ofrim, Tudor Cioara, Anghel Ionuț

Nowadays semantic models provide limited support for representing distributed AI workflows and their execution across heterogeneous edge, fog, and cloud environments. Therefore, AI processes and resources are often described using incompatible semantic representations, affecting the interoperability, orchestration, and reuse. To address these challenges, this paper proposes a SAREF-compliant ontology for representing distributed AI workflows across the edge-fog-cloud continuum. We extend the SAREF4SYST ontology with concepts for modeling AI pipelines, executable AI jobs, computational resources, deployment constraints, and communication relationships, providing a unified semantic model of both AI workflows and heterogeneous computing infrastructures. The ontology enables semantic interoperability, automated reasoning, and resource-aware orchestration of distributed AI applications while remaining fully aligned with the ETSI SAREF ecosystem. The ontology is evaluated using proof-of-concept smart grid energy services orchestration scenarios and validated using competency questions showing its ability to support AI workflow deployment, execution reasoning, and workload adaptation across heterogeneous edge, fog, and cloud environments. All competency questions were successfully validated using SPARQL querying and semantic reasoning. Experimental results demonstrate deployment success rates of 90-100% with average orchestration decision times below 80 ms across heterogeneous edge-fog-cloud environments, highlighting its effectiveness on ensuring semantic interoperability for distributed AI orchestration.

Spike-aware diffusion transformer model for energy forecasting

Authors: Andrei Mateescu, Tudor Cioara, Anghel Ionuț

Blockchain-Enabled P2P Energy Trading in Interconnected Microgrids Using Hybrid PSO-DE Optimization

Authors: Viorica Chifu, Tudor Cioara, Anghel Ionuț

Distributed JEPA: A Self-Supervised Framework for Energy Forecasting

Authors: Tudor Cioara, Anghel Ionuț

Traditional energy forecasting solutions rely on task-specific supervision and energy asset representations, limiting transferability and the ability to capture general temporal dynamics across heterogeneous assets. We address this by proposing a distributed Joint Embedding Predictive Architecture (JEPA) for self-supervised learning from heterogeneous energy time-series. The framework predicts latent representations of masked temporal segments while integrating temporal observations and contextual information within a shared embedding space. To prevent representation collapse, training combines a latent-space predictive objective with covariance and temporal variance regularization. The evaluation was conducted on energy consumption and generation datasets under data-degradation scenarios and compared with a Transformer forecasting baseline. The learned representations remained stable (cosine similarity $\approx 0.98$; effective rank 185-235). JEPA achieved performance comparable to a Transformer on building energy data, higher $R^2$ in 3/5 consumer clusters, and outperformed the baseline on 9/10 unseen PVs ($R^2$=0.73-0.88 vs. <0.45), while showing greater robustness to missing data.

Colonic Polyp Detection with Object Detection Models

Authors: Eugen Richard Ardelean, Raluca Laura Portase

In recent years, deep learning has been applied more and more to medical image analysis. One such application of deep learning is the automated polyp detection in colonoscopy with the target of reducing miss rates. This study presents a comprehensive evaluation of nine state-of-the-art object detection models for colonic polyp detection: YOLOv8, YOLOv9, YOLOv10, YOLO11, YOLO12, YOLO26, RT-DETR, YOLO-World, and YOLOE. The models were evaluated on three publicly available datasets: CVC-ClinicDB, CVC-ColonDB, and ETIS-LaribPolypDB. All models were trained under standardized conditions using identical hyperparameters and data augmentation strategies to guarantee fair comparison. Performance was evaluated using multiple metrics: mAP@50, mAP@50–95, F1 score, precision, recall, inference time, and computational cost. YOLO11 demonstrated the best overall performance, achieving mAP@50 scores of 0.995, 0.944, and 0.978 on the three datasets respectively, while maintaining the fastest inference time of approximately 150 ms per image and the third-lowest computational cost at 21.3 GFLOPs. Cross-dataset generalization experiments revealed a significant loss of performance, with mAP@50 dropping by 20–40% when models were tested on an unseen dataset, highlighting the challenge of true generalization with limited datasets. Statistical analysis by polyp size showed that while all models achieved F1 scores exceeding 0.95 for large polyps, performance decreased to 0.60–0.85 for small polyps, indicating a limitation in detecting small lesions. The analysis of failure modes showed that missed detections, false positives and boundary errors constitute 60–75% of all failures, suggesting that domain adaptation of object detection models may be required.

Object detection in histology: A multi-dataset benchmark and test-time inference

Authors: Eugen Richard Ardelean

Medical image analysis has become increasingly important for automated medical diagnosis, as well as deep learning. Specifically, object detection models may help in automatically identifying pathological structures and features. This study presents a comprehensive comparative analysis for object detection tasks in histological images of the latest models including the YOLO (You Only Look Once) architectures, from YOLOv8 to the recently introduced YOLOv12. These models were evaluated alongside alternative architectures including RT-DETR, YOLO-World, and YOLOE across five diverse histology datasets: BCNB, Nuclei, TNBC, MoNuSAC, and CryoNuSeg. The experimental analysis employed standardized training protocols with consistent hyperparameters and data augmentation strategies, evaluating the performance through multiple metrics, inference time, and computational cost. The results obtained on the five datasets indicate that YOLOv11 consistently showed a strong performance across multiple datasets, however the newly introduced attention mechanisms of YOLOv12 show good performance, despite the model having slightly lower overall performance. Specialized variants like YOLOE demonstrated promising results for specific applications, while RT-DETR showed poor performance on smaller objects, which are typical in histological images. Statistical analyses indicate that YOLOv11 indeed has the best performance but that all models have a poor performance on objects of small sizes; moreover, the most common cases of failure are background false positives and missed detections. This comprehensive evaluation provides insights for the current state of object detection architectures for clinical histopathology applications and establishes benchmarks for future avenues of research in automated medical image analysis. In addition to the multi-model benchmark, we propose Test-time Graph Similarity Propagation (TGSP), a test-time self-supervised refinement that uses ResNet50 deep features to build a k-NN similarity graph over detections and performs label propagation to re-score predicted boxes. TGSP replaces TSBP's iterative Earth-Mover matching with adaptive per-class quantile thresholds and graph-based label propagation, eliminating K-means hyperparameters and better scalability. Our analysis on histology datasets TGSP consistently matches or improves F1 relative to both a fixed 0.5 threshold and TSBP, with the biggest gains when base-model confidence calibration is poor.

Multi-Agent Coverage Using Multiplicatively Weighted Energy Voronoi Partitions

Authors: Lucian Bușoniu

We propose a coverage control method for energy-constrained multi-agent systems with single-integrator agent motion, in which agent energy discharges and recharges at constant rates. Differently existing methods, a Multiplicatively Weighted Energy Voronoi (MWEV) partition ensures that each agent’s coverage region varies with energy in such a way that it vanishes when energy drops to a reserve level sufficient to reach a charging station. We show that the optima of an energy-aware coverage objective are the dynamically evolving MWEV centroids. A generalized Lloyd algorithm provably drives agents with remaining battery to these centroids, and reserve-level agents to their charging stations, under a switched two-timescale model with fast motion and slow energy dynamics. Agents repeatedly pause coverage as they get depleted, and then resume coverage upon recharging. This happens arbitrarily many times, leading to an infinite-horizon coverage setting. The method works well in simulations, where we also apply an alternative technique with a different problem formulation. We contrast coverage performance and average agent downtime between the two methods; e.g., downtime is 17.70% of the experiment duration for our MWEV technique, compared to 28.5% for the alternative. In addition, the robustness of the proposed approach is investigated under nonlinear battery discharge dynamics, position errors, and delayed energy information.

Fast Neural-Network Approximation of Active Target Search Under Uncertainty

Authors: Lucian Bușoniu, Zsofia Lendek

We address the problem of searching for an unknown number of stationary targets at unknown positions with a mobile agent. A probability hypothesis density filter is used to estimate the expected number of targets under measurement uncertainty. Existing planners, such as Active Search (AS) and its Intermittent variant (ASI), achieve accurate detection but require costly online optimization. To reduce online computation, we propose to use a convolutional neural network to approximate AS or ASI decisions through direct inference. The network is trained on AS/ASI data using a multi-channel grid that encodes target beliefs, the agent position, visitation history, and boundary information. Simulations with uniform and clustered target distributions show that the network achieves detection rates comparable to AS or ASI while reducing computation by orders of magnitude.

Graph-based distributed Nash equilibrium seeking for potential games

Authors: Lucian Bușoniu

We introduce Distributed Asynchronous Potential function Decrease (DAPD), a discrete-time, graph-based algorithm for finding pure Nash equilibria in ordinal potential games. Two different settings are studied: one in which cost functions are twice-differentiable and convex, with Lipschitz first derivatives, and another when costs are only Lipschitz-continuous and may be non-convex. A novel graph-based update scheduler is proposed, which accelerates DAPD convergence by allowing parallel, decentralized updates of non-neighboring players. The scheduler chooses the next player to update in each neighborhood as the one with the largest decrease in cost function at the previous update. The graph topology is fixed, connected and undirected, and the update of each player depends only on its neighbors. We prove that when run with the proposed scheduler, DAPD converges to a pure Nash equilibrium in the differentiable setting, and to an ɛ -Nash equilibrium in the Lipschitz setting. In numerical experiments, DAPD with the new scheduler converges faster than with a round-robin scheduler, while better balancing the number of computations and communications than several synchronous and asynchronous baselines.

Value iteration with stopping criterion: finite iterations, stability, and near-optimality guarantees

Authors: Lucian Bușoniu

Value iteration (VI) is a cornerstone of dynamic programming that allows computing near-optimal feedback laws for general plant dynamics and cost functions. In practice, however, it must be stopped after finitely many iterations. This raises the question of when to stop the algorithm so that the resulting policies and value functions achieve desirable properties, like given near-optimality bounds and stability. In this context, we study deterministic, discrete-time systems with infinite-horizon (possibly discounted) costs whose inputs are generated by VI. We equip VI with a generalized stopping criterion that encompasses existing choices while allowing new ones. Our aim is to analyze the properties of the policies and value functions at the final iteration. Under mild assumptions, we first show that VI indeed terminates in a finite number of iterations. We then establish that the final policies are stabilizing by properly designing the stopping criterion, and derive explicit near-optimality bounds characterized by this choice. These results offer a design framework for the stopping criteria that balances computational effort with stability and performance guarantees.

Identification-Based Prediction for Event-Triggered Control in Platooning

Authors: Lucian Bușoniu

International audience

Individual and string stability regions with minimum time headway in ACC/CACC vehicular platoons

Authors: Lucian Bușoniu

Segmentation of the Retinal Vascular Network and Biomarker Quantification in OCTA Imaging

Authors: Mihnea Jurca, Raluca Didona Brehar

Forest Inspection Dataset: A Synthetic UAV Dataset for Semantic Segmentation of Forest Environments

Authors: Sergiu Nedevschi

This work describes the Forest Inspection dataset, a synthetic aerial image collection designed for semantic segmentation of forest environments with an emphasis on UAV-based forest inspection. The dataset consists of high-resolution RGB images paired with dense pixel-level semantic labels covering 11 classes, including deciduous trees, coniferous trees, fallen trees, ground vegetation, dirt ground, rocks, sky, buildings, fences, and vehicles. Images were generated in AirSim using a photorealistic virtual forest environment and captured with simulated UAV flights at three altitudes (30 m, 50 m, 80 m) and three camera pitch angles (0°, 60°, 90°) to reproduce diverse observation conditions, under two weather settings: sunny and overcast. Each data sample includes the corresponding UAV pose metadata for spatial context. The dataset is provided in standard image and annotation formats, accompanied by a description of the scene configuration and acquisition parameters. This resource is intended to support the development and evaluation of semantic segmentation models and other computer vision methods for UAV-based forest scene understanding and inspection.

Edge-Oriented Orchestration of Energy Services Using Graph-Driven Swarm Intelligence

Authors: Tudor Cioara, Vasile Ofrim, Dragos Lazea, Anca Hangan

As smart grids increasingly depend on IoT devices and distributed energy management, they require decentralized, low latency orchestration of energy services. We address this with a unified framework for edge fog cloud infrastructures tailored to smart energy systems. It features a graph based data model that captures infrastructure and workload, enabling efficient topology exploration and task placement. Leveraging this model, a swarm-based heuristic algorithm handles task offloading in a resource-aware, latency sensitive manner. Our framework ensures data interoperability via energy data space compliance and guarantees traceability using blockchain based workload notarization. We validate our approach with a real-world KubeEdge deployment, demonstrating zero downtime service migration under dynamic workloads while maintaining service continuity.

Decentralized coordination and Physics-based accountability of Grid forming inverters through AI-agent representations

Authors: Dan Mitrea, Tudor Cioara

P2P energy trading for interconnected microgrids: genetic algorithm with Hawk-Dove battery usage strategies

Authors: Tudor Cioara

Peer-to-peer energy trading among interconnected microgrids can improve local renewable energy utilization and reduce dependence on the main grid. However, existing approaches do not fully integrate participant specific battery use behavior, degradation aware pricing, energy allocation, unmatched energy reduction, and AC network feasibility. This paper proposes a genetic algorithm-based P2P trading framework for interconnected microgrids equipped with battery energy storage systems. At each trading interval, microgrids are classified as sellers, buyers, or neutral participants and assigned aggressive or conservative battery use strategies inspired by the Hawk-Dove game. Participant specific marginal energy values are derived from Lagrange multipliers and adjusted according to battery strategy and degradation cost, while a genetic algorithm determines bilaterally exchanged energy quantities by minimizing trading cost, battery degradation cost, and unmatched energy. Candidate solutions are re-paired to satisfy battery operating limits and bus-voltage and branch-loading constraints. The framework is evaluated using 5-minute renewable generation and load data for 20 microgrids connected through an IEEE 33-bus system. The results show that matched energy exceeds unmatched energy during most trading intervals, including under demand dominated conditions. Most bilateral prices provide economic benefits to both sellers and buyers, resulting in positive network level welfare on two representative days. The retained trading schedules also satisfy the considered battery and AC network constraints.

Development of a Low-Cost Measurement Platform for HF RFID Tag Antenna Performance Evaluation at 13.56 MHz

Authors: Vasile Topa, Călin Munteanu

High-frequency (HF) RFID systems operating at 13.56 MHz are widely used in applications such as near-field communication, contactless identification, and smart sensing. Their performance strongly depends on the inductive coupling between the reader and tag antennas, which is influenced by antenna geometry, relative position, orientation, and environmental conditions. This work investigates the antenna component of passive HF RFID tags, represented by planar spiral inductors, without integrating an RFID microchip, allowing the electromagnetic coupling to be analyzed independently of chip-specific effects. A low-cost automated measurement platform was developed to experimentally evaluate the influence of antenna geometry, distance, orientation, and temperature on inductively coupled HF RFID antennas. The platform combined an automated positioning system with a mobile application for remote operation, minimizing the influence of the operator during measurements. A second experimental setup was designed to investigate the effect of temperature on antenna performance. Experimental results show that rectangular spiral antennas generally provided stronger inductive coupling than the other geometries investigated. Furthermore, varying the receiving antenna orientation improved the coupling between rectangular and octagonal antennas under specific configurations. Temperature variations within the investigated range had only a minor influence on antenna performance. The proposed platform provides a low-cost, portable, and reproducible solution for the experimental characterization of HF RFID antennas operating at 13.56 MHz.

Electromagnetic Compatibility Analysis of a Battery Charger for Electric Bicycles

Authors: Vasile Topa, Călin Munteanu

As electric bicycles become a more common choice for sustainable urban mobility, their battery chargers must satisfy stringent electromagnetic compatibility (EMC) and electrical –safety requirements before being placed on the market. This paper presents an experimental EMC and electrical –safety evaluation of a battery charger for a 48 V, 250 W electric bicycle, covering electrical safety, conducted and radiated emissions, and immunity to conducted electromagnetic disturbances. The influence of a dedicated EMI filter proposed by the authors on the conducted –emission behavior of the charger was investigated by comparing measurements taken with and without the filter in place, under otherwise identical test conditions. The results confirm that the charger satisfies the applicable EMC emission and immunity limits and the electric –shock –protection aspects of electrical safety, both without and with the filter, and demonstrate that the proposed filter yields a substantial improvement in the conducted –emission compliance margin. These findings provide practical guidance for the design, verification, and certification of EMC–compliant charging equipment for electric bicycles and similar light electric vehicles.

Spatio-Temporal GraphSAGE Framework for Generalizable Power Grid Representations

Authors: Liana Maria Toderean, Tudor Cioara

Projects: AI.grids

To address the need for generalizable representation in foundation grid models, we propose a heterogeneous GraphSAGE framework that learns tokens for grid elements representation by combining temporal behaviour and physical attributes with network context obtained through edge message passing. The model is jointly trained with multiple objectives to encourage general-purpose representations, and is evaluated on multi-horizon forecasting and state estimation. Across two ACTIVSg grids, the model achieves in most cases performance comparable to task specific baselines while supporting zero-shot transfer to an unseen grid. The results suggest that the learned representations can generalize across both tasks and topologies.

Fine-Grained Complexity of Ontology Mediated Queries

Authors: Cristina Feier

Projects: HRIA - Romanian HUB for AI

Prompts and Prayers: the Rise of GPTheology

Authors: Adrian Petru Groza

Projects: HRIA - Romanian HUB for AI

Increasingly artificial intelligence (AI) has been cast in “god-like” roles (to name a few: film industry – Matrix, The Creator, Mission Impossible, Foundation, Dune etc.; literature – Children of Time, Permutation City, Neuromancer, I Have no Mouth and I Must Scream, Alphaville etc.). This trend has accelerated with the advent of sophisticated Large Language Models such as ChatGPT. For this phenomenon, where AI is perceived as divine, we use the term GPTheology, where ChatGPT and other AI models are treated as potential oracles of a semi-divine nature. This paper explores the emergence of GPTheology as a form of techno-religion, examining how narratives around AI echo traditional religious constructs. We draw on community narratives from online forums – Reddit – and recent projects – AI-powered Mazu Statue in Malaysia (Lu, 2025); “ShamAIn” Project in Korea (He-rim, 2025); AI Jesus in a Swiss Church (Kennedy, 2024). These examples show striking similarities to technological notions of the Singularity and the development of Artificial General Intelligence (AGI). Additionally, we analyse how daily interactions with AI are acquiring ritualistic associations and how AI-centric ideologies clash with or are integrated into established religions. This study uses a dataset of Reddit posts discussing AI to identify recurring themes of salvation, prophecy, and demonization surrounding AI. Our findings suggest that new belief systems are developing around AI, and this carries both philosophical and sociotechnical implications. Our paper critically analyses the benefits and dangers, as well as the social, political and ethical challenges of this development. This transdisciplinary inquiry highlights how AI and religion are increasingly intertwined, prompting necessary questions about humanity’s relationship with its creations and the future of belief.

A Comparative Survey of Social Bias in Text and Image Generation: Gaps, Directions and Compliance with the EU AI Act

Authors: Cristian Andrei Rad, Camelia Lemnaru

Generative artificial intelligence models, including large language models and image generation models, are increasingly deployed in socially impactful domains. However, these models often exhibit social biases that can amplify stereotypes and produce harmful, discriminatory outputs. In this paper, we present a modality-comparative survey of social bias in text and image generation, structured around four components: benchmarks, bias identification, measurement, and mitigation. We systematically analyze methodological parallels and divergences across the two modalities, highlighting emerging research trends and identifying gaps. Finally, we map current image generation research efforts to the EU AI Act’s technical requirements, offering insights into how the community can advance towards more fair, safe, and trustworthy systems.

MCP-Orchestrated Multi-Agent System for Automated Disinformation Detection

Authors: Adrian Petru Groza, Alexandru Lecu

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

The large spread of disinformation across digital platforms creates significant challenges to information integrity. This paper presents a multi-agent system that uses relation extraction to detect disinformation in news articles, focusing on titles and short text snippets. The proposed Agentic AI system combines four agents: (i) a machine learning agent (logistic regression), (ii) a Wikipedia knowledge check agent (which relies on named entity recognition), (iii) a coherence detection agent (using LLM prompt engineering), and (iv) a web-scraped data analyzer that extracts relational triplets for fact checking. The system is orchestrated via the Model Context Protocol (MCP), offering shared context and live learning across components. Results demonstrate that the multi-agent ensemble achieves 95.3% accuracy with an F1 score of 0.964, significantly outperforming individual agents and traditional approaches. The weighted aggregation method, mathematically derived from individual agent misclassification rates, proves superior to algorithmic threshold optimization. The modular architecture makes the system easily scalable, while also maintaining details of the decision processes.

Reducing Hallucinations in Medical AI: A Knowledge Graph-Augmented Retrieval System for Evidence-Based Age-Related Macular Degeneration Information

Authors: Alexandru Lecu, Adrian Petru Groza

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

Large language models (LLMs) have significantly advanced natural language generation but frequently produce unverified outputs, compromising their reliability in critical medical applications. We present a framework that combines structured biomedical knowledge with LLMs through retrieval-augmented generation to address this challenge. Our system automatically extracts causal relationships from 5 000 age-related macular degeneration (AMD) abstracts, building a knowledge graph with over 43 200 validated relations. Using vector-based retrieval, the framework generates contextually relevant and verifiable responses with direct clinical evidence links. We evaluated our approach across eight language models, including open-source models from 1B to 70B parameters (LLama, Mistral, Qwen, SmolLM) and GPT-5-mini, on 3 000 queries with varying question types and reasoning complexity. Smaller models (3B parameters) showed substantial improvements: SmolLM3-3B reached 95.6% accuracy on singlehop true/false questions (from 78.2% baseline). The medium-scale model Mistral-7B demonstrated the largest gains on complex multi-hop reasoning, improving from 45% to 76% accuracy on multiple-choice questions. Larger models (70B parameters) showed minimal improvement due to already high baseline performance (97-98% accuracy). Our results demonstrate that RAG-enhanced knowledge graphs enable resource-efficient smaller models to achieve performance levels approaching or matching larger models, reducing hallucinations while maintaining computational efficiency for clinical deployment [PDF](https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11298209)

TADA: Training-free Attribution and Out-of-Domain Detection of Audio Deepfakes

Authors: David Combei

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

Deepfake detection has gained significant attention across audio, text, and image modalities, with high accuracy in distinguishing real from fake. However, identifying the exact source—such as the system or model behind a deepfake—remains a less studied problem. In this paper, we take a significant step forward in audio deepfake model attribution or source tracing by proposing a training-free, green AI approach based entirely on k-Nearest Neighbors (kNN). Leveraging a pre-trained self-supervised learning (SSL) model, we show that grouping samples from the same generator is straightforward– we obtain an 0.93 F1-score across five deepfake datasets. The method also demonstrates strong out-of-domain (OOD) detection, effectively identifying samples from unseen models at an F1-score of 0.84

Unmasking real-world audio deepfakes: A data-centric approach

Authors: David Combei

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

The growing prevalence of real-world deepfakes presents a critical challenge for existing detection systems, which are often evaluated on datasets collected just for scientific purposes. To address this gap, we introduce a novel dataset of real-world audio deepfakes. Our analysis reveals that these real-world examples pose significant challenges, even for the most performant detection models. Rather than increasing model complexity or exhaustively search for a better alternative, in this work we focus on a data-centric paradigm, employing strategies like dataset curation, pruning, and augmentation to improve model robustness and generalization. Through these methods, we achieve a 55% relative reduction in EER on the In-the-Wild dataset, reaching an absolute EER of 1.7%, and a 63% reduction on our newly proposed real-world deepfakes dataset, AI4T. These results highlight the transformative potential of data-centric approaches in enhancing deepfake detection for real-world applications.

Replay Attacks Against Audio Deepfake Detection

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

We show how replay attacks undermine audio deepfake detection: By playing and re-recording deepfake audio through various speakers and microphones, we make spoofed samples appear authentic to the detection model. To study this phenomenon in more detail, we introduce ReplayDF, a dataset of recordings derived from M-AILABS and MLAAD, featuring 109 speaker-microphone combinations across six languages and four TTS models. It includes diverse acoustic conditions, some highly challenging for detection. Our analysis of six open-source detection models across five datasets reveals significant vulnerability, with the topperforming W2V2-AASIST model’s Equal Error Rate (EER) surging from 4.7% to 18.2%. Even with adaptive Room Impulse Response (RIR) retraining, performance remains compromised with an 11.0% EER. We release ReplayDF for noncommercial research use.

lujteam at SemEval-2025 Task 10: Finetuning SmolLM2 with Taxonomy-based Prompting for Explaining the Dominant Narrative in Propaganda Text

Authors: Anca Nicoleta Mărginean

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

XAI has been a long-standing goal of AI. Explaining why a text can be considered to have a dominant narrative, where the narrative is known, is of great importance for dealing with propaganda in the news. This paper reports on the participation of the system clujteam in Subtask 3 of Task 10 of Semveal 2025. The system obtained 7th place with a value of 0.72464 for F1macro, at 0.026 distance from the 1st place. The key components of the solution are the given taxonomy for the narratives and supervised fine-tuning of SmolLM2.

On the Contribution of Lexical Features to Speech Emotion Recognition

Authors: David Combei

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

Although paralinguistic cues are often considered the primary drivers of speech emotion recognition (SER), we investigate the role of lexical content extracted from speech and show that it can achieve competitive—and in some cases higher—performance compared to acoustic models. On the MELD dataset, our lexical-based approach obtains a weighted F1-score (WF1) of 51.5%, compared to 49.3% for an acousticonly pipeline with a larger parameter count. Furthermore, we analyze different self-supervised (SSL) speech and text representations, conduct a layer-wise study of transformer-based encoders, and evaluate the effect of audio denoising.

Self-Explanatory Disinformation Detection with Expert-Guided Refinement

Authors: Ioana Cheres, Adrian Petru Groza

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

To be added after acceptance

Using LLMs and ontologies to extract causal relationships from medical abstracts

Authors: Alexandru Lecu, Adrian Petru Groza

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

The substantiation of the causal relationships behind its development is very important in identifying possible interventions and early treatment. Knowledge Graphs (KG) play a crucial role in the medical research domain by organizing data into interconnected structures that represent relationships between entities such as disease, treatments, and progressions. This paper shows a complete workflow that demonstrates the extraction of causal relationships from medical abstracts using a fine-tuned GPT-based model and the integration of these relationships into a KG.

Knowledge Graph-Driven Retrieval-Augmented Generation: Integrating Deepseek-R1 with Weaviate for Advanced Chatbot Applications

Authors: Alexandru Lecu, Adrian Petru Groza

Projects: Distributed Ledger Technologies Implementation in Internal Organizational Workflows and Support Applications for Cooperation among National Security Institutions

Large language models (LLMs) have significantly advanced the field of natural language generation. However, they frequently generate unverified outputs, which compromises their reliability in critical applications. In this study, we propose an innovative framework that combines structured biomedical knowledge with LLMs through a retrieval-augmented generation technique. Our system develops a thorough knowledge graph by identifying and refining causal relationships and named entities from medical abstracts related to age-related macular degeneration (AMD). Using a vector-based retrieval process and a locally deployed language model, our framework producesresponses that are both contextually relevant and verifiable, with direct references to clinical evidence. Experimental results show that this method notably decreases hallucinations, enhancesfactual precision, and improves the clarity of generated responses, providing a robust solution for advanced biomedical chatbot applications.

Whale optimization-based clustering for buildings’ energy consumption profiles

Authors: Viorica Chifu, Tudor Cioara, Cristina Pop, Anghel Ionuț

Projects: DEDALUS

The smart grids require advanced building demand-side management solutions to balance variable renewable production. Energy consumption profile segmentation can increase the efficiency of the processes, enabling grid operators to provide more personalized requests to different customer needs and behaviors increasing engagement and participant satisfaction. However, the complexity of clustering energy profiles is amplified by the diversity of consumer behavior and the need to adapt segmentation over time as consumption patterns evolve. Establishing the number of consumer energy profile clusters in advance is challenging, as most clustering algorithms are sensitive to initial parameter settings, which can affect their performance. In this paper, we propose a new clustering method based on the Whale Optimization Algorithm (WOA), which addresses the problem of dynamicity and variability of energy data. The whale individual is defined as a set of active centroids described by a compressive set of features extracted based on peak and valley periods focused on magnitude, variation, and efficiency. To promote compact and well-separated clusters of energy profiles the Calinski-Harabasz index was used as a fitness function. Population initialization is performed using K-Means+ + algorithm which ensures a diversified initial distribution of solutions in the solution space. The evolution of solutions is ensured through the mechanisms of the WOA algorithm, which allow the gradual updating of candidate solutions to accelerate convergence and produces high-quality solutions. The efficiency of the method was evaluated on a real dataset of daily energy consumption of buildings on a university campus. The experimental results show that the WOA method outperforms the selected state-of-the-art methods for comparison. WOA obtained the lowest Davies–Bouldin index (0.527), the highest Silhouette score (0.484) and the lowest Ball–Hall coefficient (1.851), indicating superior segmentation in terms of both internal cohesion and inter-cluster separability.

A federated learning model with the whale optimization algorithm for renewable energy prediction

Authors: Viorica Chifu, Tudor Cioara, Cristina Pop, Anghel Ionuț, Liana Maria Toderean

Projects: DEDALUS

Federated prediction models for energy prosumers create a global model by combining insights from local machine learning models trained on-site without centralizing the data. For time series energy data, this collaborative approach faces challenges due to the non-IID nature of the data, variations in generation patterns, the high number of model parameters, and convergence issues, leading to poor prediction accuracy. This paper introduces a novel federated learning model, FedWOA, which uses the whale optimization algorithm to determine optimal aggregation coefficients based on the local model weight vectors by pondering the updates considering the model performance and data dimensionality construct the global shared model. To handle the non-IID data the prosumers were clustered based on the similarity of their energy profiles using K-Means. FedWOA improves the prediction quality at the prosumer site, with a 16 % average reduction of the mean absolute error compared to FedAVG while demonstrating good convergence and reduced loss.

AI-assisted electrochemical detection of bacteria siderophores in clinical samples

Authors: Adrian Petru Groza

The rapid detection of bacteria remains a significant challenge in clinical contexts, with consequences for both patients and the medical system. While the electrochemical sensors deliver fast results, the matrix effect in real samples and the overlaps of the electrochemical signals are still of major concern and cannot be overcome in the absence of biorecognition elements and samples pretreatment protocols. The implementation of portable sensors in this case has promising potential, but their use is still limited to laboratory settings. An innovative approach is to develop an AI-assisted electrochemical sensor that employs machine learning algorithms to classify the target signal and overcome the interferences from real samples with complex biological matrices. The electrochemical behavior of three bacterial markers for Pseudomonas aeruginosa , Klebsiella pneumoniae , and Escherichia coli was analyzed in human serum to estimate the impact of any potential interference, whether from the environment or between components. Chemometrics have been used to evaluate the influence that the markers have on each other's analytical signals in the case of their simultaneous detection. A Decision Tree Classifier algorithm has been applied to enhance the sensitivity and improve the accuracy of the classification in cases of complex samples with promising results on real clinical samples. • Simultaneous detection of bacteria markers was achieved via electrochemical analysis. • Chemometrics analysis identified the mutual interference effect of the molecules. • Machine learning algorithms were applied to enhance the accuracy of classification. • The training and testing data sets were done in commercial human serum.

Real-Time Electroencephalogram Data Visualization Using Generative AI Art

Authors: Adrian Petru Groza, Andrei Virgil Puiac

This study is the result of the need to research the visualization of brainwaves. The aim is based on the idea of using generative AI art systems as a method. Data visualization is an important part of understanding the evolution of the world around us. It offers the ability to see a representation that goes beyond numbers. Generative AI systems have gained the possibility of helping the process of visualizing data in new ways. This specific process includes real-time-generated artistic renderings of these data. This real-time rendering falls into the field of brainwave visualization, with the help of the EEG (electroencephalogram), which can serve here as input data for Generative AI systems. The brainwave measurement technology as a form of input to real-time generative AI systems represents a novel intersection of neuroscience and art in the field of neurofeedback art. The main question this paper hopes to address is as follows: How can brainwaves be effectively fed into generative AI art systems, and where can the outcome lead, in terms of progress? EEG data were successfully integrated with generative AI to create interactive art. The installation provided an immersive experience by moving the image with the change in the user’s mental focus, demonstrating the impact of EEG-based art.

OCTA-Based Biomarker Characterization in nAMD

Authors: Adrian Petru Groza

We aim to enhance ophthalmologists’ decision-making when diagnosing the Neovascular Age-Related Macular Degeneration (nAMD). We developed three tools to analyze Optical Coherence Tomography Angiography images: (1) extracting biomarkers such as mCNV area and vessel density using image processing; (2) generating a 3D visualization of the neovascularization for a better view of the affected regions; and (3) applying an ensemble of three white box machine learning algorithms (decision tree, support vector machines and DL-Learner) for nAMD diagnosis. The learned expressions reached 100% accuracy for the training data and 68% accuracy in testing. The main advantage is that all the learned models white-box, which ensures explainability and transparency, allowing clinicians to better understand the decision-making process.

MindBugs: Self-Explanatory Disinformation Detection System with Human-in-the-loop

Authors: Ioana Cheres, Adrian Petru Groza

Structural Retinal Analysis in Toxoplasmic Retinochoroiditis: OCT Follow-Up with Three-Dimensional Reconstruction

Authors: Adrian Petru Groza, Adrian Pop

Background: Ocular toxoplasmosis remains the leading cause of posterior uveitis worldwide. Optical coherence tomography (OCT) provides valuable insights into the structural alterations associated with this condition. The present study aimed to characterize the vitreous, retinal, and choroidal morphological changes observed during both the active and scarred stages of ocular toxoplasmosis using OCT imaging. A secondary objective was to evaluate the added value of three-dimensional reconstruction in the assessment of retinal lesions. Methods: A retrospective study was conducted on 12 eyes belonging to 12 patients diagnosed with toxoplasmosis retinochoroiditis (TRC). Optical coherence tomography (OCT) scans centered on the active lesions were qualitatively analyzed at baseline and follow-up. Additionally, a ResUNet model was trained to generate a full volumetric reconstruction of the retinochoroidal lesions in selected cases. Results: Twelve eyes were analyzed at a mean of 16.2 days from symptom onset. The mean follow-up duration was 144 days (range: 12–490 days). OCT imaging revealed characteristic alterations in the retina, choroid, and vitreous body, which were documented both at baseline and at follow-up. Representative cases were selected for three-dimensional reconstruction to illustrate the extent of retinal architectural involvement. Conclusions: OCT analysis refines our understanding of the structural damage associated with ocular toxoplasmosis, while three-dimensional reconstruction enhances our ability to visualize and interpret these alterations on a larger scale.

Explainable Fundus Image Curation and Lesion Detection in Diabetic Retinopathy

Authors: Adrian Petru Groza

Diabetic Retinopathy (DR) affects individuals with long-term diabetes. Without early diagnosis, DR can lead to vision loss. Fundus photography captures the structure of the retina along with abnormalities indicative of the stage of the disease. Artificial Intelligence (AI) can support clinicians in identifying these lesions, reducing manual workload, but models require high-quality annotated datasets. Due to the complexity of retinal structures, errors in image acquisition and lesion interpretation of manual annotators can occur. We proposed a quality-control framework, ensuring only high-standard data is used for evaluation and AI training. First, an explainable feature-based classifier is used to filter inadequate images. The features are extracted both using image processing and contrastive learning. Then, the images are enhanced and put subject to annotation, using deep-learning-based assistance. Lastly, the agreement between annotators calculated using derived formulas determines the usability of the annotations.

Towards a Common Standard for Uncertainty Quantification

Authors: Adrian Pop

Uncertainty Quantification (UQ) studies allow us todetermine whether a model is fit for a particular purpose,as well as the operational domain in which it can be used.Standardising the UQ analysis setup and result summaryenables the iterative composition of UQ information, whichis a crucial step in evaluating model credibility. In thispaper, we present an initial attempt to specify UQinformation as a cross-layer standard for Modelica-, FMI-,and SSP-based workflows subject to two essentialrestrictions: (a) uncertainties can only be described interms of parameters, and (b) analysis is limited to forwarduncertainty propagation and sensitivity analysis ofnonlinear models. More analysis features are planned forthe future. The approach is illustrated using both a simpleexample and an industrial use case.

Requirement Verification with CRML and OpenModelica

Authors: Adrian Pop

Common Requirement Modeling Language (CRML) is alanguage designed to express requirements in an intuitivemanner, in this paper we present the recent advancementsin tool support for the requirement modeling and verifica-tion workflow in OpenModelica and illustrate this on theTraffic Light use-case.

Improving ParModAuto for Equation Based Mathematical Modeling & Simulation: Metaheuristic Approach

Authors: Adrian Pop

Despite their widespread use, modeling and simulation tools still face limitations in scalability, particularly when executed in distributed computing environments. In our prior research, we introduced a task-based library named ParModAuto, designed to enable the automatic parallelization of Modelica simulation models by supporting efficient representation, clustering, scheduling, profiling, and execution of complex, dependency-heavy equation systems, which employs heuristic-driven algorithms to address computationally challenging, NP-complete optimization problems. This paper investigates whether further enhancements to the parallelization library can lead to greater speedups in simulation performance, as benchmark results indicate that additional gains are achievable.

How Open Is Open TTS? A Practical Evaluation of Open Source TTS Tools

Authors: Adriana Stan

Open-source text-to-speech (TTS) frameworks have emerged as highly adaptable platforms for developing speech synthesis systems across a wide range of languages. However, their applicability is not uniform – particularly when the target language is under-resourced or when computational resources are constrained. In this study, we systematically assess the feasibility of building novel TTS models using four widely adopted open-source architectures: FastPitch, VITS, Grad-TTS, and Matcha-TTS. Our evaluation spans multiple dimensions, including qualitative aspects such as ease of installation, dataset preparation, and hardware requirements, as well as quantitative assessments of synthesis quality for Romanian. We employ both objective metrics and subjective listening tests to evaluate intelligibility, speaker similarity, and naturalness of the generated speech. The results reveal significant challenges in tool chain setup, data preprocessing, and computational efficiency, which can hinder adoption in low-resource contexts. By grounding the analysis in reproducible protocols and accessible evaluation criteria, thiswork aims to inform best practices and promote more inclusive, language-diverse TTS development. All information needed to reproduce this study (i.e. code and data) will be made available after the final review.

ADNAC: Audio Denoiser using Neural Audio Codec

Authors: Adriana Stan

Audio denoising is critical in signal processing, enhancing intelligibility and fidelity for applications like restoring musical recordings. This paper presents a proof-of-concept for adapting a state-of-the-art neural audio codec, the Descript Audio Codec (DAC), for music denoising. This work overcomes the limitations of traditional architectures like U-Nets by training the model on a large-scale, custom-synthesized dataset built from diverse sources. Training is guided by a multi objective loss function that combines time-domain, spectral, and signal-level fidelity metrics. Ultimately, this paper aims to present a PoC for high-fidelity, generative audio restoration.

Can YOLO Detect Retinal Pathologies? A Step Towards Automated OCT Analysis

Authors: Eugen Richard Ardelean, Anca Nicoleta Mărginean

Background: Optical Coherence Tomography has become a common imaging technique that enables a non-invasive and detailed visualization of the retina and allows for the identification of various diseases. Through the advancement of technology, the volume and complexity of OCT data have rendered manual analysis infeasible, creating the need for automated means of detection. Methods: This study investigates the ability of state-of-the-art object detection models, including the latest YOLO versions (from v8 to v12), YOLO-World, YOLOE, and RT-DETR, to accurately detect pathological biomarkers in two retinal OCT datasets. The AROI dataset focuses on fluid detection in Age-related Macular Degeneration, while the OCT5k dataset contains a wide range of retinal pathologies. Results: The experiments performed show that YOLOv12 offers the best balance between detection accuracy and computational efficiency, while YOLOE manages to consistently outperform all other models across both datasets and most classes, particularly in detecting pathologies that cover a smaller area. Conclusions: This work provides a comprehensive benchmark of the capabilities of state-of-the-art object detection for medical applications, specifically for identifying retinal pathologies from OCT scans, offering insights and a starting point for the development of future automated solutions for analysis in a clinical setting.

A systematic review of transformers and large language models in the energy sector: towards agentic digital twins

Authors: Tudor Cioara, Anghel Ionuț

Projects: HRIA - Romanian HUB for AI, DEDALUS, HEDGE-IoT

Artificial intelligence (AI) has long promised to improve energy management in smart grids by enhancing situational awareness and supporting more effective decision-making. While traditional machine learning has demonstrated notable results in forecasting and optimization, it often struggles with generalization, situational awareness, and heterogeneous data integration. Recent advances in Transformer architecture, including Large Language Models (LLMs) and Foundation Models (FMs) can significantly improve the ability to model complex temporal and contextual relationships, as well as in multi-modal data fusion which is valuable for most AI applications in the energy sector. In this review, we synthesize the rapidly expanding field of AI applications in the energy domain, with a focus on Transformer Models (TMs) and LLMs, which have shown growing relevance. We examine the architectural foundations, domain-specific adaptations and practical implementations of TMs across various forecasting and grid management tasks. We then explore the emerging role of LLMs in the field: adaptation and fine tuning for the energy sector, the type of tasks they are suited for, and the new challenges they introduce. Along the way, we highlight practical implementations, innovations, and areas where the research frontier is rapidly expanding. These recent developments reviewed underscore a broader trend: Generative AI (GenAI) is beginning to augment decision-making not only in high-level planning but also in day-to-day operations, from forecasting and grid balancing to workforce training and asset onboarding. While FMs hold promises, we found limited evidence of their concrete application in energy domain to date. Therefore we introduce the concept of the Agentic Digital Twin, a next-generation model that integrates FMs to bring multi-modal situational awareness, autonomy, proactivity, and social interaction into digital twin-based energy management systems. We present the transformational impact of FMs to each phase of a digital twin and identify the open challenges that need to be addressed for their efficient and effective integration.

Hybrid transformer model with liquid neural networks and learnable encodings for buildings’ energy forecasting

Authors: Tudor Cioara, Anghel Ionuț

Projects: DEDALUS

Accurate forecasting of buildings' energy demand is essential for building operators to manage loads and resources efficiently, and for grid operators to balance local production with demand. However, nowadays models still struggle to capture nonlinear relationships influenced by external factors like weather and consumer behavior, assume constant variance in energy data over time, and often fail to model sequential data. To address these limitations, we propose a hybrid Transformer-based model with Liquid Neural Networks and learnable encodings for building energy forecasting. The model leverages Dense Layers to learn non-linear mappings to create embeddings that capture underlying patterns in time series energy data. Additionally, a Convolutional Neural Network encoder is integrated to enhance the model's ability to understand temporal dynamics through spatial mappings. To address the limitations of classic attention mechanisms, we implement a reservoir processing module using Liquid Neural Networks which introduces a controlled non-linearity through dynamic reservoir computing, enabling the model to capture complex patterns in the data. For model evaluation, we utilized both pilot data and state-of-the-art datasets to determine the model's performance across various building contexts, including large apartment and commercial buildings and small households, with and without on-site energy production. The proposed transformer model demonstrates good predictive accuracy and training time efficiency across various types of buildings and testing configurations. Specifically, SMAPE scores indicate a reduction in prediction error, with improvements ranging from 1.5 % to 50 % over basic transformer, LSTM and ANN models while the higher R² values further confirm the model's reliability in capturing energy time series variance. The 8 % improvement in training time over the basic transformer model, highlights the hybrid model computational efficiency without compromising accuracy.

Heuristic based federated learning with adaptive hyperparameter tuning for households energy prediction

Authors: Tudor Cioara, Mihai Daian, Anghel Ionuț

Projects: HEDGE-IoT, DEDALUS

Federated Learning is transforming electrical load forecasting by enabling Artificial Intelligence (AI) models to be trained directly on household edge devices. However, the prediction accuracy of federated learning models tends to diminish when dealing with non-IID data highlighting the need for adaptive hyperparameter optimization strategies to improve performance. In this paper, we propose a novel hierarchical federated learning solution for efficient model aggregation and hyperparameter tuning, specifically tailored to household energy prediction. The households with similar energy profiles are clustered at the edge, linked, and aggregated at the fog level, to enable effective and adaptive hyperparameter tuning. The federated model aggregation is optimized using hierarchical simulated annealing optimization to prioritize updates from the better-performing models. A genetic algorithm-based hyperparameter optimization method reduces the computational load on edge nodes by efficiently exploring different configurations and using only the most promising ones for edge nodes’ cross-validation. The evaluation results demonstrate a significant improvement in average prediction accuracy and better capturing of energy patterns compared to the federated averaging approach. The impact on network traffic among nodes across different layers is kept below 30 KB. Additionally, hyperparameter tuning reduces the size of model updates and the number of communication rounds by 30%, which is particularly beneficial when network resources are limited.

New care pathways for supporting transitional care from hospitals to home using AI and personalized digital assistance

Authors: Tudor Cioara, Anghel Ionuț

Projects: TransCare

Transitional care may play a vital role in the sustainability of Europe’s future healthcare system, offering solutions for relocating patient care from hospital to home, therefore addressing the growing demand for medical care as the population is ageing. Re-hospitalization for the same initial pathological cause is a recurrent problem among older adults with chronic conditions such as heart failure or chronic obstructive pulmonary disease. However, to be effective, it is nowadays essential to integrate into the transitional care process innovative Information and Communications Technology technologies to ensure that patients with comorbidities experience a smooth and coordinated transition from hospitals or care centers to home, thereby reducing the risk of rehospitalization. In this paper, we present an overview of the integration of Internet of Things, artificial intelligence, and digital assistance technologies with traditional care pathways to address the challenges and needs of healthcare systems in Europe. We identify the current gaps in transitional care and define the technology mapping to enhance the care pathways, aiming to improve patient outcomes, safety, and quality of life, avoiding hospital readmissions. Finally, we define the trial setup and evaluation methodology needed to provide clinical evidence that supports the positive impact of technology integration on patient care and discuss the potential effects on the healthcare system.

Managing Cognitive Decline Through a Social Robot–Based Intervention: Protocol for the engAGE Proof of Concept and Randomized Controlled Trial

Authors: Tudor Cioara, Anghel Ionuț

BACKGROUND: Dementia is challenging society in terms of the quality of life, the costs of health care systems, and caregivers' burden. Dementia is often preceded by a status of mild cognitive impairment (MCI), during which a healthy lifestyle and cognitive therapy seem to be effective in counteracting the decline. OBJECTIVE: The engAGE (Managing Cognitive Decline Through Theatre Therapy, Artificial Intelligence, and Social Robot-Driven Interventions) project aimed to build a technological platform to counteract cognitive decline in older adults with MCI through both cognitive therapy and lifestyle management. METHODS: The engAGE platform was built around the social robot Pepper, which provides cognitive therapy. An activity tracker and a mobile app were also integrated within the platform to help older adults with MCI monitor their sleep and physical activity, in addition to offering cognitive games at home. The proof of concept (PoC) of engAGE was designed as a 6-month long randomized controlled trial (RCT) aimed to test the solution in three European countries: Italy, Switzerland, and Norway. During this period, under the supervision of a psychologist or a therapist, Pepper provided cognitive therapy on a weekly basis at health care or daycare facilities. In parallel, the use of the mobile app and the activity tracker was recommended on a daily basis. Participants were recruited through health care institutions and care organizations and evaluated through face-to-face interviews. The primary interest of the study was to assess the impact of engAGE on cognitive capacity through the Montreal Cognitive Assessment (MoCA) and the Memory Assessment Clinic Questionnaire (MAC-Q). In addition, changes in social engagement and quality of life were measured through the University of California, Los Angeles (UCLA) Loneliness Scale and the Warwick-Edinburgh Mental Well-Being Scale (WEMWBS). The PoC also focused on the acceptability and usability of engAGE, evaluated through the System Usability Scale (SUS) and the Unified Theory of Acceptance and Use of Technology (UTAUT) questionnaire. In addition, data were collected regarding the frequency of weekly sessions and access to the mobile app, as well as activity tracker measurements. RESULTS: Data collection commenced in November 2023 and finished in July 2024, with the enrollment of 49 older adults with MCI (n=40, 81.6%, assigned to the experimental group [EG] and n=9, 18.4%, to the control group [CG]) conducted from October 2023 to April 2024. Data analysis was concluded in November 2024, and results will be published by 2026. CONCLUSIONS: The engAGE PoC represents an innovative study focused on the impact of a technology-based multidomain intervention designed for older adults with MCI that aimed to counteract cognitive decline through cognitive training, as well as improvement in terms of the quality of life, social engagement, physical activity, and sleep quality of primary end users. TRIAL REGISTRATION: ClinicalTrials.gov NCT06302686; https://clinicaltrials.gov/study/NCT06302686. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): RR1-10.2196/67601.

A Laplace diffusion-based transformer model for heart rate forecasting within daily activity context

Authors: Tudor Cioara, Anghel Ionuț, Andrei Mateescu

Projects: TransCare

With the advent of wearable Internet of Things (IoT) devices, remote patient monitoring (RPM) emerged as a promising solution for managing heart failure. However, the heart rate can fluctuate significantly due to various factors, and without correlating it to the patient's actual physical activity, it becomes difficult to assess whether changes are significant. Although Artificial Intelligence (AI) models may enhance the accuracy and contextual understanding of remote heart rate monitoring, the integration of activity data is still rarely addressed. In this paper, we propose a Transformer model combined with a Laplace diffusion technique to model heart rate fluctuations driven by physical activity of the patient. Unlike prior models that treat activity as secondary, our approach conditions the entire modeling process on activity context using specialized embeddings and attention mechanisms to prioritize activity specific historical patents. The model captures both long-term patterns and activity-specific heart rate dynamics by incorporating contextualized embeddings and dedicated encoder. The Transformer model was validated on a real-world dataset collected from 29 patients over a 4-month period. Experimental results show that our model outperforms current state-of-the-art methods, achieving a 43 % reduction in mean absolute error compared to the considered baseline models. Moreover, the coefficient of determination R2 is 0.97 indicating the model predicted heart rate is in strong agreement with actual heart rate values. These findings suggest that the proposed model can be a practical and effective tool for supporting both healthcare providers and remote patient monitoring systems.

A Data-Driven Framework for Estimating Residential Energy Flexibility for Aggregated Demand-Side Management

Authors: Tudor Cioara, Anghel Ionuț

Whale Optimization-Based Clustering for Buildings’ Energy Consumption Profiles

Authors: Tudor Cioara, Anghel Ionuț

Microgrids Coalitions for Energy Market Balancing

Authors: Tudor Cioara, Anghel Ionuț

With the integration of renewable sources in electricity distribution networks, the need to develop intelligent mechanisms for balancing the energy market has arisen. In the absence of such mechanisms, the energy market may face imbalances that can lead to power outages, financial losses or instability at the grid level. In this context, the grouping of microgrids into optimal coalitions that can absorb energy from the market during periods of surplus or supply energy to the market during periods of is a key aspect in the efficient management of distribution networks. In this article, we propose a method that identify an optimal microgrids coalition capable of addressing the dynamics of the energy market. The proposed method models the problem of identifying the optimal coalition as an optimization problem that it solves by combining a strategy inspired by cooperative game theory with a memetic algorithm. An individual is represented as a coalition of microgrids and the evolution of population of individuals over generations is assured by recombination and mutation. The fitness function is defined as the difference between the total value generated by the coalition and a penalty applied to the coalition when the energy traded by coalition exceeds the energy available/demanded on/by the energy market. The value generated by the coalition is calculated based on the profit obtained by the collation if it sells energy on the market during periods of deficit or the savings obtained by the coalition if it buys energy on the market during periods of surplus and the costs associated with the trading process. This value is divided equitably among the coalition members, according to the Shapley value, which considers the contribution of each one to the formation of collective value.

Evolutionary model for energy trading in community microgrids using Hawk-Dove strategies

Authors: Tudor Cioara, Anghel Ionuț

This paper proposes a decentralized model of energy cooperation between microgrids, in which decisions are made locally, at the level of the microgrid community. Each microgrid is modeled as an autonomous agent that adopts a Hawk or Dove strategy, depending on the level of energy stored in the battery and its role in the energy trading process. The interactions between selling and buying microgrids are modeled through an evolutionary algorithm. An individual in the algorithm population is represented as an energy trading matrix that encodes the amounts of energy traded between the selling and buying microgrids. The population evolution is achieved by recombination and mutation operators. Recombination uses a specialized operator for matrix structures, and mutation is applied to the matrix elements according to a Gaussian distribution. The evaluation of an individual is made with a multi-criteria fitness function that considers the seller profit, the degree of energy stability at the community level, penalties for energy imbalance at the community level and for the degradation of microgrids batteries. The method was tested on a simulated scenario with 100 microgrids, each with its own selling and buying thresholds, to reflect a realistic environment with variable storage characteristics of microgrids batteries. By applying the algorithm on this scenario, 95 out of the 100 microgrids reached a stable energy state. This result confirms the effectiveness of the proposed model in achieving energy balance both at the individual level, for each microgrid, and at the level of the entire community.

From Transformers to Large Language Models: A systematic review of AI applications in the energy sector towards Agentic Digital Twins

Authors: Tudor Cioara, Anghel Ionuț

Artificial intelligence (AI) has long promised to improve energy management in smart grids by enhancing situational awareness and supporting more effective decision-making. While traditional machine learning has demonstrated notable results in forecasting and optimization, it often struggles with generalization, situational awareness, and heterogeneous data integration. Recent advances in foundation models such as Transformer architecture and Large Language Models (LLMs) have demonstrated improved capabilities in modelling complex temporal and contextual relationships, as well as in multi-modal data fusion which is essential for most AI applications in the energy sector. In this review we synthesize the rapid expanding field of AI applications in the energy domain focusing on Transformers and LLMs. We examine the architectural foundations, domain-specific adaptations and practical implementations of transformer models across various forecasting and grid management tasks. We then explore the emerging role of LLMs in the field: adaptation and fine tuning for the energy sector, the type of tasks they are suited for, and the new challenges they introduce. Along the way, we highlight practical implementations, innovations, and areas where the research frontier is rapidly expanding. These recent developments reviewed underscore a broader trend: Generative AI (GenAI) is beginning to augment decision-making not only in high-level planning but also in day-to-day operations, from forecasting and grid balancing to workforce training and asset onboarding. Building on these developments, we introduce the concept of the Agentic Digital Twin, a next-generation model that integrates LLMs to bring autonomy, proactivity, and social interaction into digital twin-based energy management systems.

Technical and socio-economic perspectives for microgrid control and cooperation

Authors: Tudor Cioara, Anghel Ionuț

The microgrid offers benefits across technical, economic, environmental, and social dimensions for local energy management. However, their development faces several challenges, requiring multi- dimensional studies. In this paper, we provide a multidisciplinary overview to sustain the implementation of microgrids and to improve power supply and renewable integration. Beginning with a building-scale MG, the study will first focus on optimal techno-economic sizing to secure off-grid operation and on-site consumer specificities. The aim is then to demonstrate that resilient control solutions will ensure the effectiveness of this concept under critical conditions. We explore the deployment of these solutions at a neighborhood scale using cooperative energy management between off-grid microgrids such as virtual power plants, energy trading, and cooperative game theory. Finally, we have addressed the social dimension as it aims to assess the social factors influencing microgrid development, including a projection study on the economic growth of microgrids and several regulation guidelines.

A Study of Deep Learning Models for Audio Classification of Infant Crying in a Baby Monitoring System

Authors: Eugen Richard Ardelean

This study investigates the ability of well-known deep learning models, such as ResNet and EfficientNet, to perform audio-based infant cry detection. By comparing the performance of different machine learning algorithms, this study seeks to determine the most effective approach for the detection of infant crying, enhancing the functionality of baby monitoring systems and contributing to a more advanced understanding of audio-based deep learning applications. Understanding and accurately detecting a baby’s cries is crucial for ensuring their safety and well-being, a concern shared by new and expecting parents worldwide. Despite advancements in child health, as noted by UNICEF’s 2022 report of the lowest ever recorded child mortality rate, there is still room for technological improvement. This paper presents a comprehensive evaluation of deep learning models for infant cry detection, analyzing the performance of various architectures on spectrogram and MFCC feature representations. A key focus is the comparison between pretrained and non-pretrained models, assessing their ability to generalize across diverse audio environments. Through extensive experimentation, ResNet50 and DenseNet trained on spectrograms emerged as the most effective architectures, significantly outperforming other models in classification accuracy. Additionally, the study investigates the impact of feature extraction techniques, dataset augmentation, and model fine-tuning, providing deeper insights into the role of representation learning in audio classification. The findings contribute to the growing field of audio-based deep learning applications, offering a detailed comparative study of model architectures, feature representations, and training strategies for infant cry detection.

A Study of Deep Clustering in Spike Sorting

Authors: Eugen Richard Ardelean, Raluca Laura Portase

Spike sorting is the process of identifying the source neurons for neuronal activity recorded from extracellular electrodes. Traditional spike sorting pipelines separate the process into distinct feature extraction and clustering steps, which may not optimally capture the complex structure of spike data. This study provides a large-scale benchmark of 12 deep clustering algorithms against traditional feature extraction methods combined with K-means clustering for spike sorting. We analyze performance across 95 synthetic datasets with varying cluster counts (2-20) and complexity from the perspective of six performance metrics. Our results demonstrate that a subset of deep clustering algorithms—particularly ACeDeC, DDC, DEC, IDEC and VaDE—significantly outperform traditional methods, especially as dataset complexity increases. These deep clustering approaches effectively learn non-linear representations that better capture the structure of spike data while simultaneously optimizing clustering objectives. This dual optimization produces feature spaces tailored for clustering, combining the two traditionally separate steps of spike sorting. Our findings indicate that deep clustering approaches are most suitable for accurately identifying individual neuronal activity in extracellular recordings, providing guidance for method selection for the increasingly complex modern multi-electrode recordings.

A Study of Non-Linear Manifold Feature Extraction in Spike Sorting

Authors: Eugen Richard Ardelean, Raluca Laura Portase

With recent developments in recording hardware, the processing of neuronal data must keep up with the increasing volumes and complexity by capturing the intrinsic relationships between instances of neuronal activity while remaining invariant to noise. Here, we explore a suite of non-linear manifold feature extraction methods - including PHATE, t-SNE, UMAP, TriMap - in an attempt to identify the most adequate method for automated spike sorting. Spike sorting is the process of clustering instances of neuronal activity, called spikes, based on similarity. By embedding high-dimensional spike shapes into low-dimensional manifolds that preserve local and global structure, we demonstrate more separable and robust clusters than those obtained via traditional feature extraction methods, such as PCA. We evaluated all feature extraction methods analyzed on 95 single-channel synthetic datasets and 2 single-channel real datasets spanning a range of cluster counts. Quantitative evaluation using clustering performance metrics (such as Adjusted Rand Index, Silhouette Score, etc.) indicates that several manifold feature extractions outperform other feature extraction methods. Our results suggest that the embeddings obtained by non-linear manifold approaches can offer a powerful, high-precision option in the spike sorting of the next-generation of electrophysiological recordings. While this study focuses on single-channel data and a subset of manifold learning techniques, a baseline has been established, and future avenues of research have been opened through this work. Future work may extend these insights to multi-channel settings, such as high-density probes and incorporate emerging manifold methods, such as hierarchical and multi-view extensions, which could further improve the robustness and accuracy of spike sorting.

A reliable approach for identifying acute lymphoblastic leukemia in microscopic imaging

Authors: Lucian Bușoniu, Levente Tamás

Leukemia is a deadly disease, and the patient's recovery rate is very dependent on early diagnosis. However, its diagnosis under the microscope is tedious and time-consuming. The advancement of deep convolutional neural networks (CNNs) in image classification has enabled new techniques in automated disease detection systems. These systems serve as valuable support and secondary opinion resources for laboratory technicians and hematologists when diagnosing leukemia through microscopic examination. In this study, we deployed a pre-trained CNN model (MobileNet) that has a small size and low complexity, making it suitable for mobile applications and embedded systems. We used the L1 regularization method and a novel dataset balancing approach, which incorporates HSV color transformation, saturation elimination, Gaussian noise addition, and several established augmentation techniques, to prevent model overfitting. The proposed model attained an accuracy of 95.33% and an F1 score of 0.95 when evaluated on the held-out test set extracted from the C_NMC_2019 public dataset. We also evaluated the proposed model by adding zero-mean Gaussian noise to the test images. The experimental results indicate that the proposed model is both efficient and robust, even when subjected to additional Gaussian noise. The comparison of the proposed MobileNet_M model's results with those of ALNet and various other existing models on the same dataset underscores its superior efficacy. The code is available for reproducing the experimental results at https://tamaslevente.github.io/ALLM/.

Switched Control Strategies in Platooning With Dwell-Time-Based String Stability Guaranties

Authors: Lucian Bușoniu

This paper analyzes the behavior of a platoon of vehicles in the presence of switches between two control modes: Adaptive Cruise Control (ACC), based on local sensing, and Cooperative Adaptive Cruise Control (CACC), which incorporates vehicle-to-vehicle (V2V) communication for improved coordination. Due to the V2V communications, CACC can maintain a shorter inter-vehicular distance, resulting in better performance of the platoon. On the other hand, ACC requires a larger inter-vehicular distance but is robust to degradations introduced by the communication network (time delay, packet loss, etc.). Switching between ACC and CACC can destabilize the platoon if not carefully managed, especially due to the modification of the inter-vehicular distance. To mitigate this, we develop H∞ specifications for a class of switching systems with lower-bounded state jumps. The proposed methodology requires solving an LMI for the controller design and a minimization problem to compute a lower-bound on the minimum dwell time between two switches. The proposed approach is validated through simulations.

Learning-Based Control of the Consensus Value in Unknown Graphs

Authors: Lucian Bușoniu

We consider the problem of optimal budget allocation for consensus reaching in unknown networks. The network is represented by a directed graph whose vertices corresponds to agents influencing each other. At discrete instants, agents are influenced by an external entity that intends to sway the consensus to a target value. Between two external influence instants, the states of the agents evolve continuously due to the network dynamics. Prior results establish that, in known networks, a water-filling strategy that targets the most influential agents first is optimal. In our approach to the unknown-network setup, the marketer uses the evolution between two influence instants to update a learned model of the graph. Then, the control allocation at the next marketing instant is done according to the water-filling strategy applied to the current model of the graph. Our main analytical contribution states that the sub-optimality of the budget allocation induced by the approximation of the graph is related to the error of the learning algorithm. Extensive numerical analysis illustrates the performance of our method and suggests a regularization term that improves it further.

Multirobot path-aware global optimization

Authors: Lucian Bușoniu

We propose Voronoi Simultaneous Optimistic Optimization (VSOO), a divide-the-best-based method for multirobot global optimization of a Lipschitz-continuous physical objective (e.g., quantity of material, density of litter, signal power), whose Lipschitz constant is unknown. In this problem, a team of mobile robots must autonomously navigate as quickly as possible to all global optima of the objective function defined over their operating area. The objective can have multiple local and global optima, is initially unknown, and can only be evaluated online at robot locations. VSOO utilizes Voronoi partitions driven by the samples collected so far by the robots, which allows them to incrementally refine the search space in their simultaneous search for the optima. We guarantee everywhere-dense and global convergence for any function, and analyze convergence rates for some representative classes of function shapes. Extensive numerical simulations, performed on established classes of benchmark test functions, demonstrate that VSOO approaches all global optima faster than a series of representative source/extremum seeking techniques that – similarly to VSOO – are global optimizers designed for mobile robots. In terms of execution time, VSOO is competitive with these baselines. We finally validate VSOO in real-robot experiments in which TurtleBot3 robots successfully search for the strongest antenna signals indoors.

Consensus approximation and impulsive control for a class of uncertain multi-agent dynamics

Authors: Lucian Bușoniu

This paper studies a class of consensus dynamics where the interactions between agents are affected by a time-varying unknown scaling factor. This situation is encountered in the control of robotic fleets over a wireless network or in opinion dynamics where the confidence given to the peers varies in time. Firstly, we establish conditions under which practical upper and lower bounds on the consensus value can be determined. Secondly, we propose control strategies for allocating a given control budget to shift agent states towards a desired consensus value despite the uncertainty. We provide computationally efficient linear programming-based approaches for both problems and validate the obtained results in numerical simulations.

TanDepth: Leveraging Global DEMs for Metric Monocular Depth Estimation in UAVs

Authors: Sergiu Nedevschi

Aerial scene understanding systems face stringent payload restrictions and must often rely on monocular depth estimation for modeling scene geometry, which is an inherently ill-posed problem. Moreover, obtaining accurate ground truth data required by learning-based methods raises significant additional challenges in the aerial domain. Self-supervised approaches can bypass this problem, at the cost of providing only up-to-scale results. Similarly, recent supervised solutions which make good progress toward zero-shot generalization also provide only relative depth values. This work presents TanDepth, a practical scale recovery method for obtaining metric depth results from relative estimations at inference-time, irrespective of the type of model generating them. Tailored for uncrewed aerial vehicle (UAV) applications, our method leverages sparse measurements from Global Digital Elevation Models (GDEM) by projecting them to the camera view using extrinsic and intrinsic information. An adaptation to the cloth simulation filter is presented, which allows selecting ground points from the estimated depth map to then correlate with the projected reference points. We evaluate and compare our method against alternate scaling methods adapted for UAVs, on a variety of real-world scenes. Considering the limited availability of data for this domain, we construct and release a comprehensive, depth-focused extension to the popular UAVid dataset to further research.

Meta-heuristic federated learning aggregation methods for load forecasting

Authors: Tudor Cioara

Federated learning (FL) is essential to energy transition as it leverages decentralized energy data and machine learning to collaborative train global energy predictive models across distributed energy resources while preserving data privacy. This paper introduces one of the first FL frameworks that efficiently integrates swarm intelligence-based aggregation methods to large-scale energy consumption forecasting, by extending the TensorFlow Federated Core framework with specialized functional enhancements. The primary objective is to enhance forecasting accuracy in decentralized learning settings. We investigated the effectiveness of various nature-inspired metaheuristics for optimizing the aggregation of local model updates from distributed energy resource nodes into a global model for load forecasting tasks, including Grey Wolf Optimization (GWO), Particle Swarm Optimization (PSO), and Firefly Algorithm (FFA) against the standard Federated Averaging (FedAvg) algorithm. Using a real-world dataset comprising of 4,438 distinct energy consumers, we demonstrate that metaheuristic aggregators consistently outperform the most well-known method, Federated Averaging in predictive accuracy. Among these approaches, GWO emerges as the superior performer achieving up to 23.6% error reduction. Our findings underscore the significant potential of metaheuristic-based aggregation mechanisms in improving FL outcomes, particularly in energy forecasting applications involving large-scale distributed data scenarios.

Building Equi-Width Histograms on Homomorphically Encrypted Data

Authors: Tudor Cioara, Dragos Lazea, Anca Hangan

Histograms are widely used for summarizing data distributions, detecting anomalies, and improving machine learning models’ accuracy. However, traditional histogram-based methods require access to raw data, raising privacy concerns, particularly in sensitive IoT applications. Encryption-based techniques offer potential solutions; however, they secure the data in transit or storage, requiring decryption during analysis, which exposes raw data to potential privacy risks. In this paper, we propose a method for constructing privacy-preserving histograms directly on homomorphically encrypted IoT data, leveraging the Fast Fully Homomorphic Encryption over the Torus (TFHE) scheme implemented in the Concrete framework. To overcome the challenges posed by homomorphic encryption, we redesign the traditional histogram construction algorithm, optimizing it for secure computation by addressing constraints related to nested loops and conditional statements. As an evaluation use case, we have considered an outlier detection mechanism based on histogram frequency counts, ensuring that all data and computations remain encrypted throughout the process. Our method achieves results consistent with plaintext-based outlier detection while maintaining reasonable computational overhead compared to those reported in the existing literature.

Data Orchestration Platform for AI Workflows Execution Across Computing Continuum

Authors: Tudor Cioara

Workshop: Edge-Oriented Orchestration of Energy Services Using Graph-Driven Swarm Intelligence

Authors: Vasile Ofrim, Tudor Cioara, Dragos Lazea, Anca Hangan

Adapting the Formula for Planar Spiral Inductors’ Inductance Computation to the New Oval Geometric Shape, Ideal for Designing Wireless Power Transfer Systems for Smart Devices

Authors: Vasile Topa, Călin Munteanu

The most used spiral inductors, in the available scientific literature and in our research activities, so far, have been those with square, hexagonal, octagonal, and circular geometric shapes. Geometry plays an important role in the efficiency of these inductors when used in wireless power transfer. In this article, a new geometric shape is designed by combining the square and the circle to create an oval shape of a planar spiral inductor. Inductors with this new shape are designed, numerically modelled, and practically constructed for experimental testing. The formula for inductance computation for planar spiral inductors is adapted for this new oval shape. New geometric coefficients, required for inductance computation formula, have been determined. The new formula for inductance computation is validated both analytically, by comparing the results with those from numerical modelling, and experimentally, by comparing with measurements, for a wide range of oval spiral inductors. Five sets of different oval spiral inductors are optimally designed, numerically modelled, practically constructed, and experimentally tested. By designing this new shape for planar spiral inductors, the inductance is increased 2.16 times compared to square, 1.84 times compared to hexagonal, 2.12 times compared to octagonal, and 2.52 times compared to circular shapes. The new oval spiral inductor design will be very useful for constructing wireless power transfer systems for pacemakers, smartphones, smartwatches, and/or any other type of smart device.

Deep Clustering for Blood Cell Classification and Quantification

Authors: Adrian Petru Groza, Mihaela Macarie

Projects: HRIA - Romanian HUB for AI

Accurate classification of blood cells plays a key role in improving automated blood analysis for both medical and veterinary applications. This work presents a two-stage deep clustering method for classifying blood cells from high-dimensional signal data. In the first stage, red blood cells (RBCs) and platelets (PLTs) are separated using a combination of an improved autoencoder and the IDEC algorithm. The second stage further classifies RBC subtypes, pure RBCs, reticulocytes, and clumped RBCs, through a variational deep embedding (VaDE) approach. Due to the lack of detailed cell-level labels, soft classification probabilities are generated from sample-level data to approximate the true distributions. The aim is to contribute to the development of low-cost, automated blood analysis systems suitable for veterinary and biomedical use. Initial results indicate this method shows promise in effectively distinguishing different blood cell populations, even with limited supervision.

Where artificial intelligence stands in the development of electrochemical sensors for healthcare applications-A review

Authors: Adrian Petru Groza

The electrochemical sensor (E-sensors) market trends have identified the biomedical applications as a significant market growth with impact on personalized therapy. Given the wide variability among individuals, a key point is to acknowledge that the assays in biological samples are still limited to laboratory setup. While slight changes in the raw experimental data are beyond human capability to process, some issues related to the design of sensors identification, matrix interference, and prediction tasks can be assisted by AI tools. However, the data delivered by E-sensors for Machine Learning (ML) is not common in literature, but since the measurements can be done in real time and can identify trends and patterns, while keeping human-driven decisions in the loop, this topic is invaluable. In this work, a critical analysis of the AI-assisted sensors was performed regarding the specific tasks that can be solved by AI tools. The data flow from the design of the concept to the final results was presented related to the elaboration of E-sensors. Additionally, wearable sensors designed for biomedical applications were critically reviewed from the perspective of AI highlighting the limitations on this topic and what does the “promising” statement mean in this context.

A Qualitative Evaluation of ChatGPT4 and PaLM2’s Response to Patient’s Questions Regarding Age-Related Macular Degeneration

Authors: Adrian Petru Groza, Anca Nicoleta Mărginean

Patient compliance in chronic illnesses is essential for disease management. This also applies to age-related macular degeneration (AMD), a chronic acquired retinal degeneration that needs constant monitoring and patient cooperation. Therefore, patients with AMD can benefit by being properly informed about their disease, regardless of the condition's stage. Information is essential in keeping them compliant with lifestyle changes, regular monitoring, and treatment. Large language models have shown potential in numerous fields, including medicine, with remarkable use cases. In this paper, we wanted to assess the capacity of two large language models (LLMs), ChatGPT4 and PaLM2, to offer advice to questions frequently asked by patients with AMD. After searching on AMD-patient-dedicated websites for frequently asked questions, we curated and selected a number of 143 questions. The questions were then transformed into scenarios that were answered by ChatGPT4, PaLM2, and three ophthalmologists. Afterwards, the answers provided by the two LLMs to a set of 133 questions were evaluated by two ophthalmologists, who graded each answer on a five-point Likert scale. The models were evaluated based on six qualitative criteria: (C1) reflects clinical and scientific consensus, (C2) likelihood of possible harm, (C3) evidence of correct reasoning, (C4) evidence of correct comprehension, (C5) evidence of correct retrieval, and (C6) missing content. Out of 133 questions, ChatGPT4 received a score of five from both reviewers to 118 questions (88.72%) for C1, to 130 (97.74%) for C2, to 131 (98.50%) for C3, to 133 (100%) for C4, to 132 (99.25%) for C5, and to 122 (91.73%) for C6, while PaLM2 to 81 questions (60.90%) for C1, to 114 (85.71%) for C2, to 115 (86.47%) for C3, to 124 (93.23%) for C4, to 113 (84.97%) for C5, and to 93 (69.92%) for C6. Despite the overall high performance, there were answers that are incomplete or inaccurate, and the paper explores the type of errors produced by these LLMs. Our study reveals that ChatGPT4 and PaLM2 are valuable instruments for patient information and education; however, since there are still some limitations to these models, for proper information, they should be used in addition to the advice provided by the physicians.

Handling abort commands for household kitchen robots

Authors: Adrian Petru Groza

We propose a solution for handling abort commands given to robots. The solution is exemplified with a running scenario with household kitchen robots. The robot uses planning to find sequences of actions that must be performed in order to gracefully cancel a previously received command. The Planning Domain Definition Language (PDDL) is used to write a domain to model kitchen activities and behaviours, and this domain is enriched with knowledge from online ontologies and knowledge graphs, like DBPedia. We discuss the results obtained in different scenarios.

A novel hybrid Artificial Gorilla Troops Optimizer with Honey Badger Algorithm for solving cloud scheduling problem

Authors: Adrian Pop

Abstract Cloud computing has revolutionized the way a variety of ubiquitous computing resources are provided to users with ease and on a pay-per-usage basis. Task scheduling problem is an important challenge, which involves assigning resources to users’ Bag-of-Tasks applications in a way that maximizes either system provider or user performance or both. With the increase in system size and the number of applications, the Bag-of-Tasks scheduling (BoTS) problem becomes more complex due to the expansion of search space. Such a problem falls in the category of NP-hard optimization challenges, which are often effectively tackled by metaheuristics. However, standalone metaheuristics generally suffer from certain deficiencies which affect their searching efficiency resulting in deteriorated final performance. This paper aims to introduce an optimal hybrid metaheuristic algorithm by leveraging the strengths of both the Artificial Gorilla Troops Optimizer (GTO) and the Honey Badger Algorithm (HBA) to find an approximate scheduling solution for the BoTS problem. While the original GTO has demonstrated effectiveness since its inception, it possesses limitations, particularly in addressing composite and high-dimensional problems. To address these limitations, this paper proposes a novel approach by introducing a new updating equation inspired by the HBA, specifically designed to enhance the exploitation phase of the algorithm. Through this integration, the goal is to overcome the drawbacks of the GTO and improve its performance in solving complex optimization problems. The initial performance of the GTOHBA algorithm tested on standard CEC2017 and CEC2022 benchmarks shows significant performance improvement over the baseline metaheuristics. Later on, we applied the proposed GTOHBA on the BoTS problem using standard parallel workloads (CEA-Curie and HPC2N) to optimize makespan and energy objectives. The obtained outcomes of the proposed GTOHBA are compared to the scheduling techniques based on well-known metaheuristics under the same experimental conditions using standard statistical measures and box plots. In the case of CEA-Curie workloads, the GTOHBA produced makespan and energy consumption reduction in the range of 8.12–22.76% and 6.2–18.00%, respectively over the compared metaheuristics. Whereas for the HPC2N workloads, GTOHBA achieved 8.46–30.97% makespan reduction and 8.51–33.41% energy consumption reduction against the tested metaheuristics. In conclusion, the proposed hybrid metaheuristic algorithm provides a promising solution to the BoTS problem, that can enhance the performance and efficiency of cloud computing systems.

A Novel Artificial Electric Field Algorithm for Solving Global Optimization and Real-World Engineering Problems

Authors: Adrian Pop

The Artificial Electric Field Algorithm (AEFA) stands out as a physics-inspired metaheuristic, drawing inspiration from Coulomb’s law and electrostatic force; however, while AEFA has demonstrated efficacy, it can face challenges such as convergence issues and suboptimal solutions, especially in high-dimensional problems. To overcome these challenges, this paper introduces a modified version of AEFA, named mAEFA, which leverages the capabilities of Lévy flights, simulated annealing, and the Adaptive s-best Mutation and Natural Survivor Method (NSM) mechanisms. While Lévy flights enhance exploration potential and simulated annealing improves search exploitation, the Adaptive s-best Mutation and Natural Survivor Method (NSM) mechanisms are employed to add more diversity. The integration of these mechanisms in AEFA aims to expand its search space, enhance exploration potential, avoid local optima, and achieve improved performance, robustness, and a more equitable equilibrium between local intensification and global diversification. In this study, a comprehensive assessment of mAEFA is carried out, employing a combination of quantitative and qualitative measures, on a diverse range of 29 intricate CEC’17 constraint benchmarks that exhibit different characteristics. The practical compatibility of the proposed mAEFA is evaluated on five engineering benchmark problems derived from the civil, mechanical, and industrial engineering domains. Results from the mAEFA algorithm are compared with those from seven recently introduced metaheuristic algorithms using widely adopted statistical metrics. The mAEFA algorithm outperforms the LCA algorithm in all 29 CEC’17 test functions with 100% superiority and shows better results than SAO, GOA, CHIO, PSO, GSA, and AEFA in 96.6%, 96.6%, 93.1%, 86.2%, 82.8%, and 58.6% of test cases, respectively. In three out of five engineering design problems, mAEFA outperforms all the compared algorithms, securing second place in the remaining two problems. Results across all optimization problems highlight the effectiveness and robustness of mAEFA compared to baseline metaheuristics. The suggested enhancements in AEFA have proven effective, establishing competitiveness in diverse optimization problems.

Property probes: Live exploration of program analysis results

Authors: Adrian Pop

We present property probes, a mechanism for helping a developer explore partial program analysis results in terms of the source program interactively while the program is edited. A node locator data structure is introduced that maps between source code spans and program representation nodes, and that helps identify probed nodes in a robust way, after modifications to the source code. We have developed a client–server based tool CodeProber supporting property probes, and argue that it is very helpful in debugging and understanding program analyses. We have evaluated our tool on several languages and analyses, including a full Java compiler and a tool for intraprocedural dataflow analysis. Our performance results show that the probe overhead is negligible even when analyzing large projects.

Towards generalisable and calibrated audio deepfake detection with self-supervised representations

Authors: Adriana Stan

Generalisation—the ability of a model to perform well on unseen data—is crucial for building reliable deepfake detectors. However, recent studies have shown that the current audio deepfake models fall short of this desideratum. In this work we investigate the potential of pretrained self-supervised representations in building general and calibrated audio deepfake detection models. We show that large frozen representations coupled with a simple logistic regression classifier are extremely effective in achieving strong generalisation capabilities: compared to the RawNet2 model, this approach reduces the equal error rate from 30.9% to 8.8% on a benchmark of eight deepfake datasets, while learning less than 2k parameters. Moreover, the proposed method produces considerably more reliable predictions compared to previous approaches making it more suitable for realistic use.

TBDM-Net: Bidirectional Dense Networks with Gender Information for Speech Emotion Recognition

Authors: Adriana Stan

This paper presents a novel deep neural network-based architecture tailored for Speech Emotion Recognition (SER). The architecture capitalises on dense interconnections among multiple layers of bidirectional dilated convolutions. A linear kernel dynamically fuses the outputs of these layers to yield the final emotion class prediction. This innovative architecture is denoted as TBDM-Net: Temporally-Aware Bi-directional Dense Multi-Scale Network. We conduct a comprehensive performance evaluation of TBDM-Net, including an ablation study, across six widely-acknowledged SER datasets for unimodal speech emotion recognition. Additionally, we explore the influence of gender-informed emotion prediction by appending either golden or predicted gender labels to the architecture's inputs or predictions. The implementation of TBDM-Net is accessible at: https://github.com/adrianastan/tbdm-net

WavLM model ensemble for audio deepfake detection

Authors: David Combei, Adriana Stan

Audio deepfake detection has become a pivotal task over the last couple of years, as many recent speech synthesis and voice cloning systems generate highly realistic speech samples, thus enabling their use in malicious activities. In this paper we address the issue of audio deepfake detection as it was set in the ASVspoof5 challenge. First, we benchmark ten types of pretrained representations and show that the self-supervised representations stemming from the wav2vec2 and wavLM families perform best. Of the two, wavLM is better when restricting the pretraining data to LibriSpeech, as required by the challenge rules. To further improve performance, we finetune the wavLM model for the deepfake detection task. We extend the ASVspoof5 dataset with samples from other deepfake detection datasets and apply data augmentation. Our final challenge submission consists of a late fusion combination of four models and achieves an equal error rate of 6.56% and 17.08% on the two evaluation sets.

Efficient Training Strategies for Natural Sounding Speech Synthesis and Speaker Adaptation Based on Fastpitch

Authors: Adriana Stan

This paper focuses on adapting the functionalities of the FastPitch model to the Romanian language; extending the set of speakers from one to eighteen; synthesising speech using an anonymous identity; and replicating the identities of new, unseen speakers. During this work, the effects of various configurations and training strategies were tested and discussed, along with their advantages and weaknesses. Finally, we settled on a new configuration, built on top of the FastPitch architecture, capable of producing natural speech synthesis, for both known (identities from the training dataset) and unknown (identities learnt through short reference samples) speakers. The anonymous speaker can be used for text-to-speech synthesis, if one wants to cancel out the identity information while keeping the semantic content whole and clear. At last, we discussed possible limitations of our work, which will form the basis for future investigations and advancements.

Prompting and Learning to Detect Major Life Events from Tweets

Authors: Anca Nicoleta Mărginean

Edge Computing in Healthcare: Innovations, Opportunities, and Challenges

Authors: Tudor Cioara, Anghel Ionuț, Alexandru Rancea

Edge computing promising a vision of processing data close to its generation point, reducing latency and bandwidth usage compared with traditional cloud computing architectures, has attracted significant attention lately. The integration of edge computing in modern systems takes advantage of Internet of Things (IoT) devices and can potentially improve the systems’ performance, scalability, privacy, and security with applications in different domains. In the healthcare domain, modern IoT devices can nowadays be used to gather vital parameters and information that can be fed to edge Artificial Intelligence (AI) techniques able to offer precious insights and support to healthcare professionals. However, issues regarding data privacy and security, AI optimization, and computational offloading at the edge pose challenges to the adoption of edge AI. This paper aims to explore the current state of the art of edge AI in healthcare by using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology and analyzing more than 70 Web of Science articles. We have defined the relevant research questions, clear inclusion and exclusion criteria, and classified the research works in three main directions: privacy and security, AI-based optimization methods, and edge offloading techniques. The findings highlight the many advantages of integrating edge computing in a wide range of healthcare use cases requiring data privacy and security, near real-time decision-making, and efficient communication links, with the potential to transform future healthcare services and eHealth applications. However, further research is needed to enforce new security-preserving methods and for better orchestrating and coordinating the load in distributed and decentralized scenarios.

Demand response optimization for smart grid integrated buildings: Review of technology enablers landscape and innovation challenges

Authors: Tudor Cioara, Anghel Ionuț

Projects: DEDALUS

This paper provides a comprehensive overview and analysis of state-of-the-art technological advancements in building integration in smart grids, with a focus on enabling their participation in demand response (DR). We consolidate knowledge from high-quality sources on the main research topics, helping researchers, building owners, and energy stakeholders to stay informed about the latest developments, trends, and best practices in the field. Our review covers reputable journals papers that offer technological enablers and evidence-based insights on building interoperability, AI-based energy prediction models, demand optimization and coordination, data privacy, and decentralization. Managing buildings in DR requires careful coordination and control, thus we provide valuable insights into current practices and opportunities by examining the EU innovation projects and identifying technological innovation trends that aim to increase resident engagement by addressing regulatory and socio-economic concerns. We also discuss the main barriers to buildings’ participation in DR identifying future research directions in the field and providing mitigation insights to the building owners and grid operators. Our findings indicate that despite their potential building participation is limited due to the absence of a clear regulatory framework and lack of mature technologies to fully support and automate the programs implementation. While AI and optimization technologies show promise for improving demand coordination, challenges such as limited interoperability between buildings and energy grids, privacy concerns, and insufficient financial incentivization significantly limit the building’s participation in DR.

Whale Optimization for Cloud–Edge-Offloading Decision-Making for Smart Grid Services

Authors: Tudor Cioara, Anghel Ionuț

As IoT metering devices become increasingly prevalent, the smart energy grid encounters challenges associated with the transmission of large volumes of data affecting the latency of control services and the secure delivery of energy. Offloading computational work towards the edge is a viable option; however, effectively coordinating service execution on edge nodes presents significant challenges due to the vast search space making it difficult to identify optimal decisions within a limited timeframe. In this research paper, we utilize the whale optimization algorithm to decide and select the optimal edge nodes for executing services' computational tasks. We employ a directed acyclic graph to model dependencies among computational nodes, data network links, smart grid energy assets, and energy network organization, thereby facilitating more efficient navigation within the decision space to identify the optimal solution. The offloading decision variables are represented as a binary vector, which is evaluated using a fitness function considering round-trip time and the correlation between edge-task computational resources. To effectively explore offloading strategies and prevent convergence to suboptimal solutions, we adapt the feedback mechanisms, an inertia weight coefficient, and a nonlinear convergence factor. The evaluation results are promising, demonstrating that the proposed solution can effectively consider both energy and data network constraints while enduring faster decision-making for optimization, with notable improvements in response time and a low average execution time of approximately 0.03 s per iteration. Additionally, on complex computational infrastructures modeled, our solution shows strong features in terms of diversity, fitness evolution, and execution time.

Physics-Informed Neural Networks for Heat Pump Load Prediction

Authors: Tudor Cioara, Anghel Ionuț

Heat pumps are promising solutions for managing the increasing heating demand of residential houses, reducing the environmental impact when used with renewable energy. Accurate heat load predictions allow the heat pump to operate at the most efficient settings, maintaining comfortable temperatures while reducing excess energy use and lowering operating costs. Data-driven prediction solutions may have difficulty capturing the dynamics and nonlinearities of the thermodynamics involved. The physics-informed models combine the monitored observed data with theoretical knowledge of heat pumps and directly integrate physical constraints, allowing for better generalization and reducing the dependence on large volumes of data. However, they require detailed knowledge of the system topology and refrigerant parameters, which increases the model complexity. Therefore, in this paper, we propose a physics-informed neural network for predicting the heat load of heat pumps that integrates thermodynamics directly into the loss function of the neural network. We model the heat load as a function of the input variables, including the inlet temperature, outlet temperature, and water flow rate. We integrate the function during model training to reduce the model complexity. Our approach increases the accuracy of the predictions compared with data-driven models and generates prediction results that are consistent with the actual physical behavior of the heat pump. The results show superior prediction accuracy, with a 7.49% reduction in the RMSE and a 6.49% decrease in the MAPE, while the R2 value shows an increase of 0.02%.

An eHealth Coaching Solution to Improve Transitional Care of Seniors with Heart Failure: Long-Field Trial

Authors: Tudor Cioara, Anghel Ionuț

Heart failure is the leading reason for seniors being admitted to hospitals. Over half of the elderly individuals diagnosed with heart failure find themselves readmitted to hospitals within a span of six months. This recurrence is associated with inadequate adherence to medical treatment and recommendations, underscoring the necessity for support systems that aid seniors in better adhering to post-hospitalization instructions. The objective of this study is to evaluate the usability, usefulness and added value of the core functionalities within the H2HCare Ambient Assisted Living developed system that was evaluated with 11 participants over a long field trial of three months. Our assessment encompassed the examination of their Quality of Life as well as the usability and efficacy of the system. Overall, participants reported finding the system user-friendly, beneficial, and conducive to enhanced disease management. Improvements include tailoring the alarm system to patient standards and using a questionnaire to assess situation urgency.

Evolutionary Game for Incentivizing Social Cooperation of Prosumers in Transactive Energy Communities

Authors: Tudor Cioara, Anghel Ionuț, Dan Mitrea

Projects: DEDALUS

Transactive energy communities offer significant potential for the smart grid’s transition to renewable energy. However, they typically rely on fixed trading contracts between prosumers, which limits flexibility and engagement. Extensive research has been conducted on the mechanisms for implementing transactive energy communities, but there is still limited focus on prosumers’ trading behaviours and their social cooperation in peer-to-peer trading. In this work, we are exploring the potential of evolutionary game theory for modelling the evolution of trading behaviours among community prosumers over time aiming to define a blockchain market mechanism that promotes social cooperation-based P2P interactions. The model, inspired by the Hawk-Dove game, incentivizes prosumers’ cooperative behaviours (Doves) and penalizes prosumers with self-centred trading behaviours (Hawks), fostering social cooperation despite the temptation of higher payoffs from defection. In trading, Doves avoid conflicts and are open to cooperation, while Hawks are not. During a market session, if both prosumers choose the Hawk trading strategy, they are penalized with higher trading costs. If one chooses the Hawk and the other Dove, the Hawk gains all the benefits, and the Dove gains nothing. If both choose Dove, they share the benefits equally. We have evaluated the effectiveness of the solution with a blockchain-based energy marketplace considering realistic scenarios using energy data from prosumers. Results suggest that matched pairs of prosumers in P2P energy transactions gradually shift towards altruistic behaviours. Additionally, at the end of the market session, over 92% of the transactions involve altruistic behaviours, therefore, the prosumers are effectively incentivized to change their behaviour. Moreover, it may significantly increase traded energy in market sessions by up to 50% driven by cooperative trading behaviours of prosumers which unlock higher levels of energy flexibility. Finally, it outperforms the market clearing mechanism in transactions energy prices by up to 3%.

A Deep Q-Learning based Smart Scheduling of EVs for Demand Response in Smart Grids

Authors: Tudor Cioara, Anghel Ionuț

Economic and policy factors are driving the continuous increase in the adoption and usage of electrical vehicles (EVs). However, despite being a cleaner alternative to combustion engine vehicles, EVs have negative impacts on the lifespan of microgrid equipment and energy balance due to increased power demand and the timing of their usage. In our view grid management should leverage on EVs scheduling flexibility to support local network balancing through active participation in demand response programs. In this paper, we propose a model-free solution, leveraging Deep Q-Learning to schedule the charging and discharging activities of EVs within a microgrid to align with a target energy profile provided by the distribution system operator. We adapted the Bellman Equation to assess the value of a state based on specific rewards for EV scheduling actions and used a neural network to estimate Q-values for available actions and the epsilon-greedy algorithm to balance exploitation and exploration to meet the target energy profile. The results are promising showing that the proposed solution can effectively schedule the EVs charging and discharging actions to align with the target profile with a Person coefficient of 0.99, handling effective EVs scheduling situations that involve dynamicity given by the e-mobility features, relying only on data with no knowledge of EVs and microgrid dynamics.

Electric Vehicles Coordination for Grid Balancing Using Multi-Objective Harris Hawks Optimization

Authors: Tudor Cioara, Anghel Ionuț

Projects: DEDALUS

The rise of renewables coincides with the shift towards Electrical Vehicles (EVs) posing technical and operational challenges for the energy balance of the local grid. Nowadays, the energy grid cannot deal with a spike in EVs usage leading to a need for more coordinated and grid aware EVs charging and discharging strategies. However, coordinating power flow from multiple EVs into the grid requires sophisticated algorithms and load-balancing strategies as the complexity increases with more control variables and EVs, necessitating large optimization and decision search spaces. In this paper, we propose an EVs fleet coordination model for the day ahead aiming to ensure a reliable energy supply and maintain a stable local grid, by utilizing EVs to store surplus energy and discharge it during periods of energy deficit. The optimization problem is addressed using Harris Hawks Optimization (HHO) considering criteria related to energy grid balancing, time usage preference, and the location of EV drivers. The EVs schedules, associated with the position of individuals from the population, are adjusted through exploration and exploitation operations, and their technical and operational feasibility is ensured, while the rabbit individual is updated with a non-dominated EV schedule selected per iteration using a roulette wheel algorithm. The solution is evaluated within the framework of an e-mobility service in Terni city. The results indicate that coordinated charging and discharging of EVs not only meet balancing service requirements but also align with user preferences with minimal deviations. The assessment of the determined solutions’ quality and efficacy shows promising outcomes, with convergence after 100 iterations reflected in a generational distance of 0.35 and a Pareto front error of 1.01, while the distribution of solutions exhibits strong hypervolume thus covering a significant portion of the objective space.

Edge Offloading in Smart Grid

Authors: Tudor Cioara, Anghel Ionuț, Dragos Lazea, Anca Hangan

The energy transition supports the shift towards more sustainable energy alternatives, paving towards decentralized smart grids, where the energy is generated closer to the point of use. The decentralized smart grids foresee novel data-driven low latency applications for improving resilience and responsiveness, such as peer-to-peer energy trading, microgrid control, fault detection, or demand response. However, the traditional cloud-based smart grid architectures are unable to meet the requirements of the new emerging applications such as low latency and high-reliability thus alternative architectures such as edge, fog, or hybrid need to be adopted. Moreover, edge offloading can play a pivotal role for the next-generation smart grid AI applications because it enables the efficient utilization of computing resources and addresses the challenges of increasing data generated by IoT devices, optimizing the response time, energy consumption, and network performance. However, a comprehensive overview of the current state of research is needed to support sound decisions regarding energy-related applications offloading from cloud to fog or edge, focusing on smart grid open challenges and potential impacts. In this paper, we delve into smart grid and computational distribution architec-tures, including edge-fog-cloud models, orchestration architecture, and serverless computing, and analyze the decision-making variables and optimization algorithms to assess the efficiency of edge offloading. Finally, the work contributes to a comprehensive understanding of the edge offloading in smart grid, providing a SWOT analysis to support decision making.

Evolutionary Game for Incentivizing Social Cooperation of Prosumers in Blockchain-Based Transactive Energy Communities

Authors: Tudor Cioara, Anghel Ionuț, Dan Mitrea

D.3.1: DEDALUS technological enablers for multi-value and energy carrieragnostic residential Demand Response (1st technology release)

Authors: Anghel Ionuț

This deliverable summarises first-year project activities and their contribution to the DEDALUS value-added services. It also outlines progress and next steps across the main technological building blocks, including architecture, interoperability, ontology, flexibility estimation, blockchain, privacy, and security.

D4.1: 'First wave' of DEDALUS Digital Twins and data-driven services for residential demand response

Authors: Tudor Cioara, Anghel Ionuț

This deliverable presents the first wave of DEDALUS Digital Twins and data-driven services for residential demand response. It describes their functional and technical specifications, including architecture and algorithms, as a basis for further development and pilot deployment.

Workshop: EVs Coordination to Maximize the Usage of Local Renewable Energy

Authors: Tudor Cioara, Viorica Chifu, Anghel Ionuț

A path-based distance computation for non-convexity with applications in clustering

Authors: Eugen Richard Ardelean, Raluca Laura Portase, Rodica Potolea, Mihaela Dînșoreanu

Abstract Clustering algorithms are essential in data analysis, but evaluating their performance is challenging when the true labels are not available, especially for non-convex clusters. Traditional performance evaluation metrics struggle to identify clustering quality, often assigning higher scores for linearly separated clusters than the true clusters. We propose an original approach to distance computation that accounts for the data structure, thus improving the clustering quality evaluation for non-convex clusters without affecting other shapes of clusters. We also showcase the applicability of this method through a modified version of K-Means using the proposed method that is capable of correctly separating non-convex clusters. The validation included the analysis of performance and time complexity of 3 traditional clustering quality evaluation metrics and the K-Means clustering algorithm against their augmented versions with the proposed approach. This analysis conducted on 7 benchmark synthetic datasets and 6 real datasets with various numbers of examples and features of diverse characteristics and joint complexities: simple convex clusters, overlapped and imbalanced clusters, and non-convex clusters. Through these analyses, we show the ineffectiveness of traditional methods and that the proposed approach overcomes the weaknesses of traditional methods.

Exploration-Based Planning for Multiple-Target Search with Real-Drone Results

Authors: Lucian Bușoniu, Zsofia Lendek

Consider a drone that aims to find an unknown number of static targets at unknown positions as quickly as possible. A multi-target particle filter uses imperfect measurements of the target positions to update an intensity function that represents the expected number of targets. We propose a novel receding-horizon planner that selects the next position of the drone by maximizing an objective that combines exploration and target refinement. Confidently localized targets are saved and removed from consideration along with their future measurements. A controller with an obstacle-avoidance component is used to reach the desired waypoints. We demonstrate the performance of our approach through a series of simulations as well as via a real-robot experiment in which a Parrot Mambo drone searches from a constant altitude for targets located on the floor. Target measurements are obtained on-board the drone using segmentation in the camera image, while planning is done off-board. The sensor model is adapted to the application. Both in the simulations and in the experiments, the novel framework works better than the lawnmower and active-search baselines.

Globally convergent path-aware optimization with mobile robots

Authors: Lucian Bușoniu

Consider a mobile robot that must navigate as quickly as possible to the global maxima of a function (e.g. density of seabed litter, pollutant concentration , wireless signal strength) defined over its operating area. This objective function is initially unknown and is assumed to be Lipschitz continuous. The limited velocity of the robot restricts the next samples to neighboring positions, and to avoid wasting time and energy, the robot’s path must be adapted as new information becomes available. The paper proposes two methods that use an upper bound on the objective to iteratively change the position targeted by the robot as new samples are acquired. The first method is FTW, which Turns When the best value seen so far of the objective Function is larger than the bound of the current target position. The second is FTWD, an extension of FTW that takes into account the Distance to the target. Convergence guarantees are provided for both methods, and a convergence rate is proven to characterize how fast the FTW suboptimality decreases as the number of samples grows. In a numerical study, FTWD greatly improves performance compared to FTW, outperforms two representative source-seeking baselines, and obtains results similar to a much more computationally intensive method that does not guarantee convergence. The relationship between FTW and FTWD is also confirmed in real-robot experiments, where a TurtleBot3 seeks the darkest point on a 2D grayscale map.

Multi-Agent Active Multi-Target Search with Intermittent Measurements

Authors: Lucian Bușoniu, Zsofia Lendek

A BlueROV2-based platform for underwater mapping experiments

Authors: Lucian Bușoniu

We propose a low-cost laboratory platform for development and validation of underwater mapping techniques, using the BlueROV2 Remotely Operated Vehicle (ROV). Both the ROV and the objects to be mapped are placed in a pool that is imaged via an overhead camera. In our prototype mapping application, the ROV's pose is found using extended Kalman filtering on measurements from the overhead camera, inertial, and pressure sensors; while objects are detected with a deep neural network in the ROV camera stream. Validation experiments are performed for pose estimation, detection, and mapping. The litter detection dataset and code are made publicly available.

Control design for robust platoons with communication delays

Authors: Lucian Bușoniu

International audience

A novel method for landslide deformation monitoring by fusing UAV photogrammetry and LiDAR data based on each sensor's mapping advantage in regards to terrain feature

Authors: Sergiu Nedevschi

Landslides are global hazards that contribute significantly to worldwide catastrophes. Since landslides cause fatalities and property damage, understanding movement patterns is crucial to mitigate risks or potential reactivations. Slope kinematic modeling can utilize geodetic surveying or direct observations, though Unmanned Aerial Vehicles with photogrammetric or laser scanning technologies have become prevalent in recent decades. This study introduces a new approach for monitoring landslides through UAV photogrammetry and LiDAR data. LiDAR and photogrammetry raster images were integrated by merging them according to landscape features, encompassing areas with vegetation and human-made structures. This fusion method leverages the complementary strengths of photogrammetry and LiDAR, optimizing their capabilities for distinct terrain features: photogrammetry excels in capturing detailed textures, while LiDAR penetrates vegetation to provide accurate ground-level data. By integrating these technologies, the Digital Feature Model (DFM) offers superior accuracy compared to conventional elevation models, making it particularly effective for monitoring complex terrains. To assess the robustness of the proposed methodology, 496 points were measured with topo-geodetic instruments within an active landslide, each point representing different terrain features. This analysis demonstrated the effective merging of data from both sensors, leading to thorough results that accurately represented the ground surface with the most suitable technology. The RMSE values for the vertical differences between the ground truth and the proposed DFM were 0.060 m, notably lower than the 0.206 m from photogrammetry and 0.441 m from LiDAR. This novel method can enhance the evaluation of landslide dynamics and movements, as follow-up surveys using this approach yield greater accuracy regarding differences between epochs. • Topic of high interest for the Geoscience and Soil Science academic community. • State of the art instrumentation used for hazard mapping and monitoring. • Innovative method enhances the accuracy of established methods. • Novel approach to data fusion between photogrammetry and LiDAR datasets. • Possible methodological improvements and future use in other research studies.

Semantic Segmentation of Remote Sensing Images With Transformer-Based U-Net and Guided Focal-Axial Attention

Authors: Sergiu Nedevschi

In the field of remote sensing, semantic segmentation of Unmanned Aerial Vehicle (UAV) imagery is crucial for tasks such as land resource management, urban planning, precision agriculture, and economic assessment. Traditional methods use Convolutional Neural Networks (CNNs) for hierarchical feature extraction but are limited by their local receptive fields, restricting comprehensive contextual understanding. To overcome these limitations, we propose a combination of transformer and attention mechanisms to improve object classification, leveraging their superior information modeling capabilities to enhance scene understanding. In this paper, we present SwinFAN (Swin-based Focal Axial attention Network), a U-Net framework featuring a Swin transformer as encoder, equipped with a novel decoder that introduces two new components for enhanced semantic segmentation of urban remote sensing images. The first proposed component is a Guided Focal-Axial (GFA) attention module which combines local and global contextual information, enhancing the model's ability to discern intricate details and complex structures. The second component is an innovative Attentionbased Feature Refinement Head (AFRH) designed to improve the precision and clarity of segmentation outputs through selfattention and convolutional techniques. Comprehensive experiments demonstrate that the accuracy of our proposed architecture significantly outperforms state-of-the-art models. More specifically, our method achieves mIoU (mean Intersection over Union) improvements of 1.9% on UAVid, 3.6% on Potsdam, 1.9% on Vaihingen, and 0.8% on LoveDA.

Improving Counting Accuracy of Post-Disaster Visual Question Answering for Remote Sensing

Authors: Sergiu Nedevschi

In post-disaster damage assessment, Visual Question Answering (VQA) systems are essential in identifying the severity and scope of damage. However, counting-related tasks, such as determining the number of vehicles and flooded buildings, remain a significant challenge for current deep learning models. To address this issue, we propose DeVANet (DeBERTa Vision Attention Network), a novel architecture aimed at enhancing counting accuracy in VQA for post-disaster scenarios. We leverage DeBERTa for language modeling and introduce an innovative image embedding module, where local-global attention guides Vision Mamba features to achieve precise extraction of both small and large object features. Our fusion mechanism employs self-attention for both text and image features, followed by bi-directional cross-attention and co-attention to enhance multimodal integration. We tackle VQA as both a classification and regression problem by employing separate MLPs for each task: one handling discrete class predictions and the other generating continuous values for counting tasks. A joint loss function, combining weighted cross-entropy and negative binomial loss, ensures optimized performance across both tasks. Extensive experiments on the FloodNet and RescueNet datasets demonstrate that DeVANet achieves significant improvements in counting accuracy and overall VQA performance compared to state-of-the-art works, supported by detailed ablation studies that validate the effectiveness of each component in the architecture.

Forest Inspection Dataset - Sunny sequences

Authors: Sergiu Nedevschi

Humans use UAVs to monitor changes in forest environments since they are lightweight and provide a large variety of surveillance data. However, their information does not present enough details for understanding the scene which is needed to assess the degree of deforestation. Deep learning algorithms must be trained on large amounts of data to output accurate interpretations, but ground truth recordings of annotated forest imagery are not available. To solve this problem, we introduce a new large aerial dataset for forest inspection which contains both real-world and virtual recordings of natural environments, with densely annotated semantic segmentation labels and depth maps, taken in different illumination conditions, at various altitudes and recording angles. We test the performance of two multi-scale neural networks for solving the semantic segmentation task (HRNet and PointFlow network), studying the impact of the various acquisition conditions and the capabilities of transfer learning from virtual to real data. Our results showcase that the best results are obtained when the training is done on a dataset containing a large variety of scenarios, rather than separating the data into specific categories. We also develop a framework to assess the deforestation degree of an area.

Recent advances on energy management and control of direct current microgrid for smart cities and industry: A Survey

Authors: Tudor Cioara

Recently, the intent to use microgrid (MG) technology for urban, residential, and industrial applications has significantly increased. Thanks to the integration of shared storage technologies, these power systems allow for higher penetration of distributed renewable generation (DGs) and better mitigation of imbalances between demand and generation. This responds to social and environmental requirements in terms of decarbonizing energy and also contributes to strengthening smart cities. DC (Direct Current) microgrids offer several advantages compared to AC (Alternating Current) type microgrids, like superior efficiency, better control, stability, compatibility with the DC nature of renewables and storage sources, and the absence of reactive and synchronization problems. However, before fully exploiting the potential of microgrids in renewable-powered smart grids, it is necessary to conduct further research and discussion on critical technical and socio-economic challenges. This paper presents a review of the existing state-of-the-art research in DC microgrid development, relevant challenges related to security, communication, power quality, and operation, as well as the appropriate control and energy management strategies to handle them. As control and energy management strategies considerably impact other performance indicators such as operating cost, emissions, and power system safety, this paper offers a perspective on the potential improvement of such management solutions.

Reinforcement-Learning-Based Edge Offloading Orchestration in Computing Continuum

Authors: Tudor Cioara

The AI-driven applications and large data generated by IoT devices connected to large-scale utility infrastructures pose significant operational challenges, including increased latency, communication overhead, and computational imbalances. Addressing these is essential to shift the workloads from the cloud to the edge and across the entire computing continuum. However, to achieve this, significant challenges must still be addressed, particularly in decision making to manage the trade-offs associated with workload offloading. In this paper, we propose a task-offloading solution using Reinforcement Learning (RL) to dynamically balance workloads and reduce overloads. We have chosen the Deep Q-Learning algorithm and adapted it to our workload offloading problem. The reward system considers the node’s computational state and type to increase the utilization of the computational resources while minimizing latency and bandwidth utilization. A knowledge graph model of the computing continuum infrastructure is used to address environment modeling challenges and facilitate RL. The learning agent’s performance was evaluated using different hyperparameter configurations and varying episode lengths or knowledge graph model sizes. Results show that for a better learning experience, a low, steady learning rate and a large buffer size are important. Additionally, it offers strong convergence features, with relevant workload tasks and node pairs identified after each learning episode. It also demonstrates good scalability, as the number of offloading pairs and actions increases with the size of the knowledge graph and the episode count.

Privacy Preserving Anomaly Detection on Homomorphic Encrypted Data from IoT Sensors

Authors: Tudor Cioara, Anca Hangan, Dragos Lazea

IoT devices have become indispensable components of our lives, and the advancement of AI technologies will make them even more pervasive, increasing the vulnerability to malfunctions or cyberattacks and raising privacy concerns. Encryption can mitigate these challenges; however, most existing anomaly detection techniques decrypt the data to perform the analysis, potentially undermining the encryption protection provided during transit or storage. Homomorphic encryption schemes are promising solutions as they enable the processing and execution of operations on IoT data while still encrypted, however, these schemes offer only limited operations, which poses challenges to their practical usage. In this paper, we propose a novel privacy-preserving anomaly detection solution designed for homomorphically encrypted data generated by IoT devices that efficiently detects abnormal values without performing decryption. We have adapted the Histogram-based anomaly detection technique for TFHE scheme to address limitations related to the input size and the depth of computation by implementing vectorized support operations. These operations include addition, value placement in buckets, labeling abnormal buckets based on a threshold frequency, labeling abnormal values based on their range, and bucket labels. Evaluation results show that the solution effectively detects anomalies without requiring data decryption and achieves consistent results comparable to the mechanism operating on plain data. Also, it shows robustness and resilience against various challenges commonly encountered in IoT environments, such as noisy sensor data, adversarial attacks, communication failures, and device malfunctions. Moreover, the time and computational overheads determined for several solution configurations, despite being large, are reasonable compared to those reported in existing literature.

Workshop: Customizing Pre-Processing Algorithms for Streaming Sensor Data on Embedded Networked Devices

Authors: Tudor Cioara, Dragos Lazea, Anca Hangan

Burst detection in neuronal activity

Authors: Eugen Richard Ardelean

Neurons in the brain fire action potentials or spikes to encode information and communicate with each other. The firing can be in tonic mode, where individual spikes are emitted at relatively distant time intervals, and in burst mode, where neurons typically fire trains of spikes in short temporal succession. Functionally, neuronal bursts may be involved in reliable information transmission or specialized information encoding schemes. Understanding the dynamics of neuronal burst activity plays a crucial role in unravelling the complex mechanisms underlying neural information processing. However, detecting bursts in extracellularly recorded neural data is far from trivial. There are many methods for burst detection, but none was adopted universally, and they are based exclusively on the evaluation of spike times. This article aims to compare these burst detection algorithms against each other from several perspectives in order to establish the most suitable and robust one. Moreover, we propose a novel burst detection method considering additional features that contain as much information as possible from the recording.

Spike sorting using Superlets: Identifying feature importance through perturbation

Authors: Eugen Richard Ardelean

Spike sorting is a technique in the field of neuroscience applied to identify and classify the action potentials (spikes) of neurons in extracellular recordings. This method is important considering that the simultaneous activity of several neurons can be recorded through the use of electrodes, but the recorded signals are often mixed. Extracting the characteristics of individual spikes provides valuable information about the identity of neurons that generate them, the type of cells involved (inhibitory/excitatory), encoding of information in spike rates or spike times, and so on. Therefore, unmixing the spikes recorded on an electrode via spike sorting is crucial. Here, we apply a novel method to extract features of mixed spikes, namely the Superlet transform, which enables the computation of spectral characteristics with a higher resolution. We use machine learning to determine which features of the Superlet spectrum contain the most information about the shapes of individual spikes, thereby enabling their unmixing during spike sorting.

Symbolic Analysis Based Pipeline for EEG Data

Authors: Eugen Richard Ardelean

This paper is concerned with developing an analysis tool for electroencephalogram (EEG) data. It explores the possibility of using the self-organizing map algorithm as starting point in a symbolic analysis based approach, with the aim of reducing the challenges brought by the high dimensionality and complexity of EEG data. The solution represents a pipeline integrating different steps for accomplishing an in-depth analysis: self-organizing map training, clustering, color sequence generation, pattern specificity index, pattern triggered average and peristimulus time histogram computation.

Self-sufficient energy communities using cooperative games over blockchain and smart contracts

Authors: Tudor Cioara, Anghel Ionuț, Viorica Chifu, Cristina Pop

The small-scale prosumers' decentralized cooperation and aggregation in energy communities are essential to increase renewable energy penetration and to ensure a successful energy transition. Despite their potential small-scale prosumers are not motivated to participate in local energy value chains due to a lack of trust and decentralized cooperation models for capturing local sustainability preferences, most innovation efforts are focused on financial incentives that are anyway very low. In this paper, we propose a solution for prosumers’ decentralized coordination using a cooperative game on top of a blockchain overlay by considering their complementary energy-related features and flexibility for self-sufficient energy communities. We define a governance model based on cooperative games and blockchain-sharing mechanisms reinforce the prosumers' trust to support the decentralized self-organization in coalitions to balance the renewable generation and demand. Self-enforcing smart contracts are used for the blockchain integration of the governance model to allow the decentralized creation and management of prosumers’ coalitions for optimized payoff distribution towards self-sufficiency and using tokens capturing the beyond-financial community level exchanges. The evaluation results show our solution's effectiveness in enabling prosumers to cooperate and create self-sufficient coalitions in a decentralized fashion with low gas consumption and transactional time overhead.

The Predictive Capabilities of Artificial Intelligence-Based OCT Analysis for Age-Related Macular Degeneration Progression—A Systematic Review

Authors: Adrian Petru Groza, Anca Nicoleta Mărginean

The era of artificial intelligence (AI) has revolutionized our daily lives and AI has become a powerful force that is gradually transforming the field of medicine. Ophthalmology sits at the forefront of this transformation thanks to the effortless acquisition of an abundance of imaging modalities. There has been tremendous work in the field of AI for retinal diseases, with age-related macular degeneration being at the top of the most studied conditions. The purpose of the current systematic review was to identify and evaluate, in terms of strengths and limitations, the articles that apply AI to optical coherence tomography (OCT) images in order to predict the future evolution of age-related macular degeneration (AMD) during its natural history and after treatment in terms of OCT morphological structure and visual function. After a thorough search through seven databases up to 1 January 2022 using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, 1800 records were identified. After screening, 48 articles were selected for full-text retrieval and 19 articles were finally included. From these 19 articles, 4 articles concentrated on predicting the anti-VEGF requirement in neovascular AMD (nAMD), 4 articles focused on predicting anti-VEGF efficacy in nAMD patients, 3 articles predicted the conversion from early or intermediate AMD (iAMD) to nAMD, 1 article predicted the conversion from iAMD to geographic atrophy (GA), 1 article predicted the conversion from iAMD to both nAMD and GA, 3 articles predicted the future growth of GA and 3 articles predicted the future outcome for visual acuity (VA) after anti-VEGF treatment in nAMD patients. Since using AI methods to predict future changes in AMD is only in its initial phase, a systematic review provides the opportunity of setting the context of previous work in this area and can present a starting point for future research.

Artificial Intelligence for Personalised Ophthalmology Residency Training

Authors: Adrian Petru Groza, Anca Nicoleta Mărginean, Radu-Răzvan Slăvescu

Residency training in medicine lays the foundation for future medical doctors. In real-world settings, training centers face challenges in trying to create balanced residency programs, with cases encountered by residents not always being fairly distributed among them. In recent years, there has been a tremendous advancement in developing artificial intelligence (AI)-based algorithms with human expert guidance for medical imaging segmentation, classification, and prediction. In this paper, we turned our attention from training machines to letting them train us and developed an AI framework for personalised case-based ophthalmology residency training. The framework is built on two components: (1) a deep learning (DL) model and (2) an expert-system-powered case allocation algorithm. The DL model is trained on publicly available datasets by means of contrastive learning and can classify retinal diseases from color fundus photographs (CFPs). Patients visiting the retina clinic will have a CFP performed and afterward, the image will be interpreted by the DL model, which will give a presumptive diagnosis. This diagnosis is then passed to a case allocation algorithm which selects the resident who would most benefit from the specific case, based on their case history and performance. At the end of each case, the attending expert physician assesses the resident's performance based on standardised examination files, and the results are immediately updated in their portfolio. Our approach provides a structure for future precision medical education in ophthalmology.

Ontology engineering with Large Language Models

Authors: Adrian Petru Groza

We tackle the task of enriching ontologies by automatically translating natural language sentences into Description Logic. Since Large Language Models (LLMs) are the best tools for translations, we fine-tuned a GPT-3 model to convert Natural Language sentences into OWL Functional Syntax. We employ objective and concise examples to fine-tune the model regarding: instances, class subsumption, domain and range of relations, object properties relationships, disjoint classes, complements, cardinality restrictions. The resulted axioms are used to enrich an ontology, in a human supervised manner. The developed tool is publicly provided as a Protge plugin.

Brave new world: Artificial Intelligence in teaching and learning

Authors: Adrian Petru Groza, Anca Nicoleta Mărginean

We exemplify how Large Language Models are used in both teaching and learning. We also discuss the AI incidents that have already occurred in the education domain, and we argue for the urgent need to introduce AI policies in universities and for the ongoing strategies to regulate AI. Regarding policy for AI, our view is that each institution should have a policy for AI in teaching and learning. This is important from at least twofolds: (i) to raise awareness on the numerous educational tools that can both positively and negatively affect education; (ii) to minimise the risk of AI incidents in education.

Measuring reasoning capabilities of ChatGPT

Authors: Adrian Petru Groza

I shall quantify the logical faults generated by ChatGPT when applied to reasoning tasks. For experiments, I use the 144 puzzles from the library \url{https://users.utcluj.ro/~agroza/puzzles/maloga}~\cite{groza:fol}. The library contains puzzles of various types, including arithmetic puzzles, logical equations, Sudoku-like puzzles, zebra-like puzzles, truth-telling puzzles, grid puzzles, strange numbers, or self-reference puzzles. The correct solutions for these puzzles were checked using the theorem prover Prover9~\cite{mccune2005release} and the finite models finder Mace4~\cite{mccune2003mace4} based on human-modelling in Equational First Order Logic. A first output of this study is the benchmark of 100 logical puzzles. For this dataset ChatGPT provided both correct answer and justification for 7\% only. %, while BARD for 5\%. Since the dataset seems challenging, the researchers are invited to test the dataset on more advanced or tuned models than ChatGPT3.5 with more crafted prompts. A second output is the classification of reasoning faults conveyed by ChatGPT. This classification forms a basis for a taxonomy of reasoning faults generated by large language models. I have identified 67 such logical faults, among which: inconsistencies, implication does not hold, unsupported claim, lack of commonsense, wrong justification. The 100 solutions generated by ChatGPT contain 698 logical faults. That is on average, 7 fallacies for each reasoning task. A third ouput is the annotated answers of the ChatGPT with the corresponding logical faults. Each wrong statement within the ChatGPT answer was manually annotated, aiming to quantify the amount of faulty text generated by the language model. On average, 26.03\% from the generated text was a logical fault.

Case Study: Using AI-Assisted Code Generation In Mobile Teams

Authors: Adrian Petru Groza

The aim of this study is to evaluate the performance of AI-assisted programming in actual mobile development teams that are focused on native mobile languages like Kotlin and Swift. The extensive case study involves 16 participants and 2 technical reviewers, from a software development department designed to understand the impact of using LLMs trained for code generation in specific phases of the team, more specifically, technical onboarding and technical stack switch. The study uses technical problems dedicated to each phase and requests solutions from the participants with and without using AI-Code generators. It measures time, correctness, and technical integration using ReviewerScore, a metric specific to the paper and extracted from actual industry standards, the code reviewers of merge requests. The output is converted and analyzed together with feedback from the participants in an attempt to determine if using AI-assisted programming tools will have an impact on getting developers onboard in a project or helping them with a smooth transition between the two native development environments of mobile development, Android and iOS. The study was performed between May and June 2023 with members of the mobile department of a software development company based in Cluj-Napoca, with Romanian ownership and management.

Solving probability puzzles with logic toolkit

Authors: Adrian Petru Groza

The proposed approach is to formalise the probabilistic puzzle in equational FOL. Two formalisations are needed: one theory for all models of the given puzzle, and a second theory for the favorable models. Then Mace4 - that computes all the interpretation models of a FOL theory - is called twice. First, it is asked to compute all the possible models M p .Second, the additional constraint is added, and Mace4 computes only favourabile models M f. Finally, the definition of probability is applied: the number of favorable models is divided by the number of possible models. The proposed approach equips students from the logic tribe to find the correct solution for puzzles from the probabilitistic tribe, by using their favourite instruments: modelling and formalisation. I have exemplified here five probabilistic puzzles and how they can be solved by translating the min FOL and then find the corresponding interpretation models. Mace4 was the tool of choice here. Ongoing work is investigating the limits of this method on various collections of probabilistic puzzles

Detecting diabetic retinopathy severity through fundus images using an ensemble of classifiers

Authors: Adrian Petru Groza

Diabetic retinopathy is an ocular condition that affects individuals with diabetes mellitus. It is a common complication of diabetes that can impact the eyes and lead to vision loss. One method for diagnosing diabetic retinopathy is the examination of the fundus of the eye. An ophthalmologist examines the back part of the eye, including the retina, optic nerve, and the blood vessels that supply the retina. In the case of diabetic retinopathy, the blood vessels in the retina deteriorate and can lead to bleeding, swelling, and other changes that affect vision. We proposed a method for detecting diabetic diabetic severity levels. First, a set of data-prerpocessing is applied to available data: adaptive equalisation, color normalisation, Gaussian filter, removal of the optic disc and blood vessels. Second, we perform image segmentation for relevant markers and extract features from the fundus images. Third, we apply an ensemble of classifiers and we assess the trust in the system.

Interleaving GANs with knowledge graphs to support design creativity for book covers

Authors: Adrian Petru Groza

An attractive book cover is important for the success of a book. In this paper, we apply Generative Adversarial Networks (GANs) to the book covers domain, using different methods for training in order to obtain better generated images. We interleave GANs with knowledge graphs to alter the input title to obtain multiple possible options for any given title, which are then used as an augmented input to the generator. Finally, we use the discriminator obtained during the training phase to select the best images generated with new titles. Our method performed better at generating book covers than previous attempts, and the knowledge graph gives better options to the book author or editor compared to using GANs alone.

Formalising Natural Language Quantifiers for Human-Robot Interactions

Authors: Adrian Petru Groza

We present a method for formalising quantifiers in natural language in the context of human-robot interactions. The solution is based on first-order logic extended with capabilities to represent the cardinality of variables, operating similarly to generalised quantifiers. To demonstrate the method, we designed an end-to-end system able to receive input as natural language, convert it into a formal logical representation, evaluate it, and return a result or send a command to a simulated robot.

Forest Mixing: Investigating the Impact of Multiple Search Trees and a Shared Refinements Pool on Ontology Learning

Authors: Adrian Petru Groza

We aim at development white-box machine learning algorithms. We focus here on algorithms for learning axioms in description logic. We extend the Class Expression Learning for Ontology Engineering (CELOE) algorithm contained in the DL-Learner tool. The approach uses multiple search trees and a shared pool of refinements in order to split the search space in smaller subspaces. We introduce the conjunction operation of best class expressions from each tree, keeping the results which give the most information. The aim is to foster exploration from a diverse set of starting classes and to streamline the process of finding class expressions in ontologies. The current implementation and settings indicated that the Forest Mixing approach did not outperform the traditional CELOE. Despite these results, the conceptual proposal brought forward by this approach may stimulate future improvements in class expression finding in ontologies.

The Common Requirement Modeling Language

Authors: Adrian Pop

CRML (the Common Requirement Modeling Language) is a new language for the formal expression of requirements. The ambition is to release the language as an open standard integrated into the open source modeling and simulation tool OpenModelica and interoperable with the open systems engineering standard SysMLv2. CRML allows to express requirements as multidisciplinary spatiotemporal constraints that can be verified against system design by co-simulating requirements models with behavioral models. Particular attention is paid to the following aspects. The requirements models must be easily legible and sharable between disciplines and stakeholders and must capture realistic constraints on the system, including time-dependent constraints with probabilistic criteria, in recognition of the fact that no constraint can be fulfilled at any time at any cost. The theoretical foundation of the language lies on 4-valued Boolean algebra, set theory and function theory. The coupling of the requirements models to the behavioral models is obtained through the specification of bindings, the automatic generation of Modelica code from the CRML model and use of the FMI and SSP standards. CRML and the proposed methodology is compatible with SysMLv2, forming a comprehensive work-flow and tool-chain encompassing requirement analysis, system design and V&V. The final objective is to facilitate the demonstration of correctness of system behavior against assumptions and requirements by building a workflow around Model-Driven Engineering and Open Standards for automating the creation of verification simulators.

A Semi-Automatic Translator from System Dynamics to Modelica with Application to Socio-Bio-Physical Systems

Authors: Adrian Pop

System Dynamics is a modelling paradigm that has been applied to a wide range of systems, from economic to physical and from managerial to ecological. The main strength of the paradigm is its ease of use. A System Dynamics modeller does not need to focus on equations; instead, models are expressed in terms of stocks and flows. Modelica, on the other hand, is an equation-based modelling language capable of multi-domain modelling using equations. It gives the user more freedom but requires more mathematical focus and skills. Therefore, a unification of equation-based modelling and the System Dynamics paradigm is seen as highly beneficial. Advantages include the ability for System Dynamics modellers to use the tools available in the Modelica ecosystem. Furthermore, it allows the integration of System Dynamics models into Modelica models. To achieve this goal, we developed an XMILE-to-Modelica translator that maps System Dynamics models represented in the XMILE standard exchange format to Modelica models. We also applied a Modelica-to-Julia translator to demonstrate the generality of the approach.We translated several models to test the correctness of the translator. In particular, the Earth System Climate Interpretable Model (ESCIMO) was translated from its original version in the Vensim toolkit into the OpenModelica toolkit, and a correct validation was obtained by comparing simulation results between simulators. Our work improves tool interoperability and further demonstrates the feasibility of using Modelica as a unified, standard language to integrate models created using System Dynamics, including large and complex socio-biophysical systems.

Towards generalisable and calibrated synthetic speech detection with self-supervised representations

Authors: Adriana Stan

Generalisation -- the ability of a model to perform well on unseen data -- is crucial for building reliable deepfake detectors. However, recent studies have shown that the current audio deepfake models fall short of this desideratum. In this work we investigate the potential of pretrained self-supervised representations in building general and calibrated audio deepfake detection models. We show that large frozen representations coupled with a simple logistic regression classifier are extremely effective in achieving strong generalisation capabilities: compared to the RawNet2 model, this approach reduces the equal error rate from 30.9% to 8.8% on a benchmark of eight deepfake datasets, while learning less than 2k parameters. Moreover, the proposed method produces considerably more reliable predictions compared to previous approaches making it more suitable for realistic use.

An analysis on the effects of speaker embedding choice in non auto-regressive TTS

Authors: Adriana Stan

In this paper we introduce a first attempt on understanding how a non-autoregressive factorised multi-speaker speech synthesis architecture exploits the information present in different speaker embedding sets. We analyse if jointly learning the representations, and initialising them from pretrained models determine any quality improvements for target speaker identities. In a separate analysis, we investigate how the different sets of embeddings impact the network’s core speech abstraction (i.e.zero conditioned) in terms of speaker identity and representation learning. We show that, regardless of the used set of embeddings and learning strategy, the network can handle various speaker identities equally well, with barely noticeable variations in speech output quality, and that speaker leakage within the core structure of the synthesis system is inevitable in the standard training procedures adopted thus far.

An analysis of large speech models-based representations for speech emotion recognition

Authors: Adriana Stan

Large speech models-derived features have recently shown increased performance over signal-based features across multiple downstream tasks, even when the networks are not finetuned towards the target task. In this paper we show the results of an analysis of several signal- and neural models-derived features for speech emotion recognition. We use pretrained models and explore their inherent potential abstractions of emotions. Simple classification methods are used so as to not interfere or add knowledge to the task. We show that, even without finetuning, some of these large neural speech models' representations can enclose information that enables performances close to, and even beyond state-of-the-art results across six standard speech emotion recognition datasets.

Diabetic Macular Edema Optical Coherence Tomography Biomarkers Detected with EfficientNetV2B1 and ConvNeXt

Authors: Anca Nicoleta Mărginean

(1) Background: Diabetes mellitus (DM) is a growing challenge, both for patients and physicians, in order to control the impact on health and prevent complications. Millions of patients with diabetes require medical attention, which generates problems regarding the limited time for screening but also addressability difficulties for consultation and management. As a result, screening programs for vision-threatening complications due to DM have to be more efficient in the future in order to cope with such a great healthcare burden. Diabetic macular edema (DME) is a severe complication of DM that can be prevented if it is timely screened with the help of optical coherence tomography (OCT) devices. Newly developing state-of-the-art artificial intelligence (AI) algorithms can assist physicians in analyzing large datasets and flag potential risks. By using AI algorithms in order to process OCT images of large populations, the screening capacity and speed can be increased so that patients can be timely treated. This quick response gives the physicians a chance to intervene and prevent disability. (2) Methods: This study evaluated ConvNeXt and EfficientNet architectures in correctly identifying DME patterns on real-life OCT images for screening purposes. (3) Results: Firstly, we obtained models that differentiate between diabetic retinopathy (DR) and healthy scans with an accuracy of 0.98. Secondly, we obtained a model that can indicate the presence of edema, detachment of the subfoveolar neurosensory retina, and hyperreflective foci (HF) without using pixel level annotation. Lastly, we analyzed the extent to which the pretrained weights on natural images "understand" OCT scans. (4) Conclusions: Pretrained networks such as ConvNeXt or EfficientNet correctly identify features relevant to the differentiation between healthy retinas and DR, even though they were pretrained on natural images. Another important aspect of our research is that the differentiation between biomarkers and their localization can be obtained even without pixel-level annotation. The "three biomarkers model" is able to identify obvious subfoveal neurosensory detachments, retinal edema, and hyperreflective foci, as well as very small subfoveal detachments. In conclusion, our study points out the possible usefulness of AI-assisted diagnosis of DME for lowering healthcare costs, increasing the quality of life of patients with diabetes, and reducing the waiting time until an appropriate ophthalmological consultation and treatment can be performed.

A Machine Learning Enhanced Approach for Automated Sunquake Detection in Acoustic Emission Maps

Authors: Anca Nicoleta Mărginean

Abstract Sunquakes are seismic emissions visible on the solar surface, associated with some solar flares. Although discovered in 1998, they have only recently become a more commonly detected phenomenon. Despite the availability of several manual detection guidelines, to our knowledge, the astrophysical data produced for sunquakes is new to the field of machine learning. Detecting sunquakes is a daunting task for human operators, and this work aims to ease and, if possible, to improve their detection. Thus, we introduce a dataset constructed from acoustic egression-power maps of solar active regions obtained for Solar Cycles 23 and 24 using the holography method. We then present a pedagogical approach to the application of machine-learning representation methods for sunquake detection using autoencoders, contrastive learning, object detection and recurrent techniques, which we enhance by introducing several custom, domain-specific data augmentation transformations. We address the main challenges of the automated sunquake-detection task, namely the very high noise patterns in and outside the active region shadow and the extreme class imbalance given by the limited number of frames that present sunquake signatures. With our trained models, we find temporal and spatial locations of peculiar acoustic emission and qualitatively associate them to eruptive and high energy emission. While noting that these models are still in a prototype stage, and there is much room for improvement in metrics and bias levels, we hypothesize that their agreement on example use cases has the potential to enable detection of weak solar acoustic manifestations.

Privacy-Preserving Computation for Peer-to-Peer Energy Trading on a Public Blockchain

Authors: Tudor Cioara, Anghel Ionuț, Dan Mitrea

To ensure the success of energy transition and achieve the target of reducing the carbon footprint of energy systems, the management of energy systems needs to be decentralized. Public blockchains offer favorable features to support energy sector democratization and reinforce citizens' trust, such as tamper-proof energy data registration and sharing, decentralization, transparency, and support for peer-to-peer (P2P) energy trading. However, in blockchain-based P2P energy markets, transactional data are public and accessible, which raises privacy concerns related to prosumers' energy profiles while lacking scalability and featuring high transactional costs. In this paper, we employ secure multi-party computation (MPC) to assure privacy on a P2P energy flexibility market implementation in Ethereum by combining the prosumers' flexibility orders data and storing it safely on the chain. We provide an encoding mechanism for orders on the energy market to obfuscate the amount of energy traded by creating groups of prosumers, by splitting the amount of energy from bids and offers, and by creating group-level orders. The solution wraps around the smart contracts-based implementation of an energy flexibility marketplace, assuring privacy features on all market operations such as order submission, matching bids and offers, and commitment in trading and settlement. The experimental results show that the proposed solution is effective in supporting P2P energy flexibility trading, reducing the number of transactions, and gas consumption with a limited computational time overhead.

Cooperative Games Over Blockchain and Smart Contracts for Self-Sufficient Energy Communities

Authors: Tudor Cioara, Anghel Ionuț

The household prosumers’ decentralized cooperation and aggregation in energy communities are essential to increase renewable energy penetration and to ensure a successful energy transition. Despite their potential, the prosumers are not motivated to participate in local energy value chains due to the lack of trust and decentralized cooperation models for meeting community welfare and sustainability preferences, most innovation efforts being focused on financial incentives that are anyway very low. In this paper, we propose a solution for prosumers’ decentralized coordination in self-sufficient energy communities using cooperative games on top of a blockchain overlay that considers their complementary energy features and flexibility mobilization. The proposed model for community-level local energy balance fits well in circumstances in which there is a strong motivation in the community to prioritize sustainability and environmental concerns and reduce dependence on external energy. We define a governance model to support the decentralized self-organization of prosumers in coalitions for balancing the renewable generation and demand while considering via tokenization factors that go beyond purely economic motivations and foster cooperation and collaboration among the prosumers in the community. Self-enforcing contracts are used to implement the cooperative game model enabling the decentralized management of prosumers’ coalitions for optimized tokens-based payoff distribution towards self-sufficiency. The evaluation results show our solution’s effectiveness in facilitating the prosumers cooperation in self-sufficient coalitions achieving a minimal difference between the energy consumption and production in the community of approximately 0.01%, with a low transactional time overhead.

Review of Blockchain Tokens Creation and Valuation

Authors: Tudor Cioara, Anghel Ionuț, Dan Mitrea

Blockchain and tokens are relatively new research areas insufficiently explored from both technical and economic perspectives. Even though tokens provide benefits such as easier market access, increased liquidity, lower transaction costs, and automated transactional process, their valuation and price determination are still challenging due to factors such as a lack of intrinsic value, volatility, and regulation making trading risky. In this paper, we address this knowledge gap by reviewing the existing literature on token creation and valuation to identify and document the factors affecting their valuation, investment, and founding, as well as the most promising domains of applicability. The study follows the PRISMA methodology and uses the Web of Science database, defining clear research questions and objective inclusion criteria for the articles. We discuss token technical development, including creating, issuing, and managing tokens on an Ethereum blockchain using smart contracts. The study revealed several key factors that significantly impact the field of tokenomics: demand and supply, social incentives, market conditions, macroeconomics, collective behavior, speculation, and inclusion in index funds. The most relevant use cases of blockchain and tokens are related to the digitization of virtual and physical assets, accountability, and traceability usual in smart grids or supply chains management, social governance, and art and gamification including metaverse.

Blockchain Solution for Buildings’ Multi-Energy Flexibility Trading Using Multi-Token Standards

Authors: Tudor Cioara, Anghel Ionuț

Buildings can become a significant contributor to an energy system’s resilience if they are operated in a coordinated manner to exploit their flexibility in multi-carrier energy networks. However, research and innovation activities are focused on single-carrier optimization (i.e., electricity), aiming to achieve Zero Energy Buildings, and miss the significant flexibility that buildings may offer through multi-energy coupling. In this paper, we propose to use blockchain technology and ERC-1155 tokens to digitize the heat and electrical energy flexibility of buildings, transforming them into active flexibility assets within integrated multi-energy grids, allowing them to trade both heat and electricity within community-level marketplaces. The solution increases the level of interoperability and integration of the buildings with community multi-energy grids and brings advantages from a transactive perspective. It permits digitizing multi-carrier energy using the same token and a single transaction to transfer both types of energy, processing transaction batches between the sender and receiver addresses, and holding both fungible and non-fungible tokens in smart contracts to support energy markets’ financial payments and energy transactions’ settlement. The results show the potential of our solution to support buildings in trading heat and electricity flexibility in the same market session, increasing their interoperability with energy markets while decreasing the transactional overhead and gas consumption.

Graph Learning and Deep Neural Network Ensemble for Supporting Cognitive Decline Assessment

Authors: Tudor Cioara, Anghel Ionuț, Alexandru Rancea

Cognitive decline represents a significant public health concern due to its severe implications on memory and general health. Early detection is crucial to initiate timely interventions and improve patient outcomes. However, traditional diagnosis methods often rely on personal interpretations or biases, may not detect the early stages of cognitive decline, or involve invasive screening procedures; thus, there is a growing interest in developing non-invasive methods benefiting also from the technological advances. Wearable devices and Internet of Things sensors can monitor various aspects of daily life together with health parameters and can provide valuable data regarding people’s behavior. In this paper, we propose a technical solution that can be useful for potentially supporting cognitive decline assessment in early stages, by employing advanced machine learning techniques for detecting higher activity fragmentation based on daily activity monitoring using wearable devices. Our approach also considers data coming from wellbeing assessment questionnaires that can offer other important insights about a monitored person. We use deep neural network models to capture complex, non-linear relationships in the daily activities data and graph learning for the structural wellbeing information in the questionnaire answers. The proposed solution is evaluated in a simulated environment on a large synthetic dataset, the results showing that our approach can offer an alternative as a support for early detection of cognitive decline during patient-assessment processes.

A Machine Learning Pipeline to Forecast the Electricity and Heat Consumption in a City District

Authors: Tudor Cioara, Anghel Ionuț

The shift towards renewable energy integration into smart grids has led to complex management processes, which require finer-grained energy and heat generation/ demand forecasting while considering data from monitoring devices and the integration of smaller multi-energy sub-systems at the community, district, or buildings level. However, energy prediction is challenging due to the high variability in the electrical and thermal energy demands of building occupants, the heterogenous characteristics of the energy assets or buildings in a district, and the length of the forecasting horizon. In this paper, we define a data-driven machine-learning pipeline to predict the electricity and thermal consumption of buildings and energy assets from a city district in 24 h intervals. Each pipeline’s step is divided into sensors’ data processing and model integration, data enrichment and features engineering, and multilayer perceptron model training. To address some of the drawbacks of using the multi-layer perceptron model, such as slow convergence rate and risk of overfitting, and to ensure a lower error in the energy prediction process features, an engineering technique was employed. We incorporated weather data features and interaction features derived from fusing the energy data with statistical models to capture the nonlinear patterns of the electrical and heat demands. The proposed approach was successfully validated in a real-world environment, a city district in Gent, Belgium. It featured good prediction results for electricity and heat production and consumption of various assets without considering the physical characteristics, making it viable and easily applicable in broader urban areas. The evaluation of energy prediction accuracy yielded good results, with a Mean Absolute Error (MAE) falling within the range of 0.003 to 3.27, and a Mean Absolute Scaled Error (MASE) ranging from 7 × 10−5 to 2.57 × 10−3.

FedWOA: A Federated Learning Model that uses the Whale Optimization Algorithm for Renewable Energy Prediction

Authors: Tudor Cioara, Anghel Ionuț

Privacy is important when dealing with sensitive personal information in machine learning models, which require large data sets for training. In the energy field, access to household prosumer energy data is crucial for energy predictions to support energy grid management and large-scale adoption of renewables however citizens are often hesitant to grant access to cloud-based machine learning models. Federated learning has been proposed as a solution to privacy challenges however report issues in generating the global prediction model due to data heterogeneity, variations in generation patterns, and the high number of parameters leading to even lower prediction accuracy. This paper addresses these challenges by introducing FedWOA a novel federated learning model that employs the Whale Optimization Algorithm to aggregate global prediction models from the weights of local LTSM neural network models trained on prosumer energy data. The proposed solution identifies the optimal vector of weights in the search spaces of the local models to construct the global shared model and then is subsequently transmitted to the local nodes to improve the prediction quality at the prosumer site while for handling non-IID data K-Means was used for clustering prosumers with similar scale of energy data. The evaluation results on prosumers energy data have shown that FedWOA can effectively enhance the accuracy of energy prediction models accuracy by 25% for MSE and 16% for MAE compared to FedAVG while demonstrating good convergence and reduced loss.

Smart Contracts and Homomorphic Encryption for Private P2P Energy Trading and Demand Response on Blockchain

Authors: Tudor Cioara, Anghel Ionuț, Dan Mitrea

Social Factors in P2P Energy Trading Using Hedonic Games

Authors: Tudor Cioara, Anghel Ionuț, Dan Mitrea

Lately, the energy communities have gained a lot of attention as they have the potential to significantly contribute to the resilience and flexibility of the energy system, facilitating widespread integration of intermittent renewable energy sources. Within these communities the prosumers can engage in peer-to-peer trading, fostering local collaborations and increasing awareness about energy usage and flexible consumption. However, even under these favorable conditions, prosumer engagement levels remain low, requiring trading mechanisms that are aligned with their social values and expectations. In this paper, we introduce an innovative hedonic game coordination and cooperation model for P2P energy trading among prosumers which considers the social relationships within an energy community to create energy coalitions and facilitate energy transactions among them. We defined a heuristic that optimizes the prosumers coalitions, considering their social and energy price preferences and balancing the energy demand and supply within the community. We integrated the proposed hedonic game model into a state-of-the-art blockchain-based P2P energy flexibility market and evaluated its performance within an energy community of prosumers. The evaluation results on a blockchain-based P2P energy flexibility market show the effectiveness in considering social factors when creating coalitions, increasing the total amount of energy transacted in a market session by 5% compared with other game theory-based solutions. Finally, it shows the importance of the social dimensions of P2P energy transactions, the positive social dynamics in the energy community increasing the amount of energy transacted by more than 10% while contributing to a more balanced energy demand and supply within the community.

Fedwoa: A Federated Learning Model with the Whale Optimization Algorithm for Renewable Energy Prediction

Authors: Tudor Cioara, Anghel Ionuț

Small field dosimetry overview

Authors: Eugen Barbulescu

of the charge transfer in an ionized complex of three stacked guanines .....

A study of autoencoders as a feature extraction technique for spike sorting

Authors: Eugen Richard Ardelean, Mihaela Dînșoreanu

Spike sorting is the process of grouping spikes of distinct neurons into their respective clusters. Most frequently, this grouping is performed by relying on the similarity of features extracted from spike shapes. In spite of recent developments, current methods have yet to achieve satisfactory performance and many investigators favour sorting manually, even though it is an intensive undertaking that requires prolonged allotments of time. To automate the process, a diverse array of machine learning techniques has been applied. The performance of these techniques depends however critically on the feature extraction step. Here, we propose deep learning using autoencoders as a feature extraction method and evaluate extensively the performance of multiple designs. The models presented are evaluated on publicly available synthetic and real "in vivo" datasets, with various numbers of clusters. The proposed methods indicate a higher performance for the process of spike sorting when compared to other state-of-the-art techniques.

Sharp detection of oscillation packets in rich time-frequency representations of neural signals

Authors: Eugen Richard Ardelean

Brain oscillations most often occur in bursts, called oscillation packets, which span a finite extent in time and frequency. Recent studies have shown that these packets portray a much more dynamic picture of synchronization and transient communication between sites than previously thought. To understand their nature and statistical properties, techniques are needed to objectively detect oscillation packets and to quantify their temporal and frequency extent, as well as their magnitude. There are various methods to detect bursts of oscillations. The simplest ones divide the signal into band limited sub-components, quantifying the strength of the resulting components. These methods cannot by themselves cope with broadband transients that look like genuine oscillations when restricted to a narrow band. The most successful detection methods rely on time-frequency representations, which can readily show broadband transients and harmonics. However, the performance of such methods is conditioned by the ability of the representation to localize packets simultaneously in time and frequency, and by the capabilities of packet detection techniques, whose current state of the art is limited to extraction of bounding boxes. Here, we focus on the second problem, introducing two detection methods that use concepts derived from clustering and topographic prominence. These methods are able to delineate the packets' precise contour in the time-frequency plane. We validate the new approaches using both synthetic and real data recorded in humans and animals and rely on a super-resolution time-frequency representation, namely the superlets, as input to the detection algorithms. In addition, we define robust tests for benchmarking and compare the new methods to previous techniques. Results indicate that the two methods we introduce shine in low signal-to-noise ratio conditions, where they only miss a fraction of packets undetected by previous methods. Finally, algorithms that delineate precisely the border of spectral features and their subcomponents offer far more valuable information than simple rectangular bounding boxes (time and frequency span) and can provide a solid foundation to investigate neural oscillations' dynamics.

Improved space breakdown method – A robust clustering technique for spike sorting

Authors: Eugen Richard Ardelean, Mihaela Dînșoreanu

Space Breakdown Method (SBM) is a clustering algorithm that was developed specifically for low-dimensional neuronal spike sorting. Cluster overlap and imbalance are common characteristics of neuronal data that produce difficulties for clustering methods. SBM is able to identify overlapping clusters through its design of cluster centre identification and the expansion of these centres. SBM's approach is to divide the distribution of values of each feature into chunks of equal size. In each of these chunks, the number of points is counted and based on this number the centres of clusters are found and expanded. SBM has been shown to be a contender for other well-known clustering algorithms especially for the particular case of two dimensions while being too computationally expensive for high-dimensional data. Here, we present two main improvements to the original algorithm in order to increase its ability to deal with high-dimensional data while preserving its performance: the initial array structure was substituted with a graph structure and the number of partitions has been made feature-dependent, denominating this improved version as the Improved Space Breakdown Method (ISBM). In addition, we propose a clustering validation metric that does not punish overclustering and such obtains more suitable evaluations of clustering for spike sorting. Extracellular data recorded from the brain is unlabelled, therefore we have chosen simulated neural data, to which we have the ground truth, to evaluate more accurately the performance. Evaluations conducted on synthetic data indicate that the proposed improvements reduce the space and time complexity of the original algorithm, while simultaneously leading to an increased performance on neural data when compared with other state-of-the-art algorithms. Code available at: https://github.com/ArdeleanRichard/Space-Breakdown-Method.

Active search and coverage using point-cloud reinforcement learning

Authors: Lucian Bușoniu

We consider a problem in which the trajectory of a mobile 3D sensor must be optimized so that certain objects are both found in the overall scene and covered by the point cloud, as fast as possible. This problem is called target search and coverage, and the paper provides an end-to-end deep reinforcement learning (RL) solution to solve it. The deep neural network combines four components: deep hierarchical feature learning occurs in the first stage, followed by multi-head transformers in the second, max-pooling and merging with bypassed information to preserve spatial relationships in the third, and a distributional dueling network in the last stage. To evaluate the method, a simulator is developed where cylinders must be found by a Kinect sensor. A network architecture study shows that deep hierarchical feature learning works for RL and that by using farthest point sampling (FPS) we can reduce the amount of points and achieve not only a reduction of the network size but also better results. We also show that multi-head attention for point-clouds helps to learn the agent faster but converges to the same outcome. Finally, we compare RL using the best network with a greedy baseline that maximizes immediate rewards and requires for that purpose an oracle that predicts the next observation. We decided RL achieves significantly better and more robust results than the greedy strategy.

Optimistic planning for control of hybrid-input nonlinear systems

Authors: Lucian Bușoniu

We propose two branch-and-bound, optimistic planning algorithms for discrete-time nonlinear optimal control problems in which there is a continuous and a discrete action (input). The dynamics and rewards (negative costs) must be Lipschitz but can otherwise be general, as long as certain boundedness conditions are satisfied by the continuous action, reward, and Lipschitz constant of the dynamics. We start by investigating the structure of the space of hybrid-input sequences. Based on this structure, we propose for the first algorithm an optimistic selection rule that picks for refinement (branching) the subset with the largest upper bound on the value. At the price of a higher budget, the second method reduces the reliance on the Lipschitz constant, by refining all sets that are potentially optimistic. This effectively means that the Lipschitz constant is automatically optimized. The way to select the largest-impact action along which to refine the sets is the same for both algorithms, and still depends on the Lipschitz constant. We provide convergence rate guarantees for both methods, which link the computational budget to the near-optimality of the action sequences returned, in a way that depends on a problem complexity measure. We also give empirical results for a nonlinear problem, where the algorithms are applied in receding horizon, and depending on the budget either one or the other algorithm works better.

Multi-Agent Exploration-Based Search for an Unknown Number of Targets

Authors: Lucian Bușoniu, Zsofia Lendek

This paper presents an active sensor fusion technique for multiple mobile agents (robots) to detect an unknown number of static targets at unknown positions. To process and fuse sensor measurements from the agents, we use a random finite set formulation with an iterated-corrector probability hypothesis density filter. Our main contribution is to introduce two different multi-agent planners to quickly find the targets. The planners make greedy decisions for the next state of each agent by maximizing an objective function consisting of target refinement and exploration components. We demonstrate the performance of our approach through a series of simulations using homogeneous and heterogeneous agents. The results show that our framework works better than a lawnmower baseline, and that a centralized version of the planner works best.

Near-optimal control of nonlinear systems with hybrid inputs and dwell-time constraints

Authors: Lucian Bușoniu

We propose two new optimistic planning algorithms for nonlinear hybrid-input systems, in which the input has both a continuous and a discrete component, and the discrete component must respect a dwell-time constraint. Both algorithms select sets of input sequences for refinement at each step, along with a continuous or discrete step to refine (split). The dwell-time constraint means that the discrete splits must keep the discrete mode constant if the required dwell-time is not yet reached. Convergence rate guarantees are provided for both algorithms, which show the dependency between the near-optimality of the sequence returned and the computational budget. The rates depend on a novel complexity measure of the dwell-time constrained problem. We present simulation results for two problems, an adaptive-quantization networked control system and a model for the COVID pandemic.

3D exploration-based search for multiple targets using a UAV

Authors: Lucian Bușoniu, Zsofia Lendek

Consider an unmanned aerial vehicle (UAV) that searches for an unknown number of targets at unknown positions in 3D space. A particle filter uses imperfect measurements about the targets to update an intensity function that represents the expected number of targets. We propose a receding-horizon planner that selects the next UAV position by maximizing a joint, exploration and target-refinement objective. Confidently localized targets are saved and removed from consideration. A nonlinear controller with an obstacle-avoidance component is used to reach the desired waypoints. We demonstrate the performance of our approach through a series of simulations, as well as in real-robot experiments with a Parrot Mambo drone that searches for targets from a constant altitude. The proposed planner works better than a lawnmower and a target-refinement-only method.

Two-Channel Extended Kalman Filtering with Intermittent Measurements

Authors: Lucian Bușoniu, Zsofia Lendek, Ștefan Pîrje

We consider two nonlinear state estimation problems in a setting where an extended Kalman filter receives measurements from two sets of sensors via two channels (2C). In the stochastic-2C problem, the channels drop measurements stochastically, whereas in 2C scheduling, the estimator chooses when to read each channel. In the first problem, we generalize linear-case 2C analysis to obtain -- for a given pair of channel arrival rates -- boundedness conditions for the trace of the error covariance, as well as a worst-case upper bound. For scheduling, an optimization problem is solved to find arrival rates that balance low channel usage with low trace bounds, and channels are read deterministically with the expected periods corresponding to these arrival rates. We validate both solutions in simulations for linear and nonlinear dynamics; as well as in a real experiment with an underwater robot whose position is being intermittently found in a UAV camera image.

Hepatocellular Carcinoma Recognition from Ultrasound Images Using Combinations of Conventional and Deep Learning Techniques

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea, Raluca Didona Brehar

Hepatocellular Carcinoma (HCC) is the most frequent malignant liver tumor and the third cause of cancer-related deaths worldwide. For many years, the golden standard for HCC diagnosis has been the needle biopsy, which is invasive and carries risks. Computerized methods are due to achieve a noninvasive, accurate HCC detection process based on medical images. We developed image analysis and recognition methods to perform automatic and computer-aided diagnosis of HCC. Conventional approaches that combined advanced texture analysis, mainly based on Generalized Co-occurrence Matrices (GCM) with traditional classifiers, as well as deep learning approaches based on Convolutional Neural Networks (CNN) and Stacked Denoising Autoencoders (SAE), were involved in our research. The best accuracy of 91% was achieved for B-mode ultrasound images through CNN by our research group. In this work, we combined the classical approaches with CNN techniques, within B-mode ultrasound images. The combination was performed at the classifier level. The CNN features obtained at the output of various convolution layers were combined with powerful textural features, then supervised classifiers were employed. The experiments were conducted on two datasets, acquired with different ultrasound machines. The best performance, above 98%, overpassed our previous results, as well as representative state-of-the-art results.

The use of budget UAV systems and GIS spatial analysis in cadastral and construction surveying for building planning

Authors: Sergiu Nedevschi

The uncertainty that comes with planning, constructing, and maintaining buildings is a constant issue for architects and civil engineers. As topography is the framework that unites architecture and landscape, the design and planning projects heavily rely on a range of monitoring, surveying methods and comprehensive field data. Along with the traditional topo-geodetic instrumentation used in land and construction surveying, unmanned aerial vehicles equipped with digital cameras and structure from motion software have been increasingly used recently in a variety of fields to create high-resolution digital elevation models. Despite this widespread use, in the majority of surveying projects it is considered that the topographic representations produced through this technology is inferior to that obtained with surveys conducted using conventional methods, along with other constraints imposed by legislation, environment and weather conditions. While certain limitations of unmanned aerial vehicle (UAV) systems are challenging, their advantage for gathering data from a different perspective and the generated outputs have the potential to significantly advance the construction industry. The present article provides an overview of the usefulness of budget UAV systems in developing a methodology that accompanies the conventional survey process for civil engineering applications. Thus, along with the established survey for cadastral and technical documentations necessary for the architectural process, a complementary UAV survey was developed, with subsequent spatial analysis in a geographic information system (GIS), in order to expand the array of deliverables. These include useful orthophoto map, larger-scale and denser representations of the topography, digital surface and terrain models, slope, aspect and solar radiation maps which will offer helpful information and instructions at the start of the construction planning process. The methodology contains two case studies with different degrees of terrain and vegetation challenges, and also presents an accuracy assessment and overall benefits discussion regarding the UAV implementation.

Deep Learning Techniques for Liver Tumor Recognition in Ultrasound Images

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea

Cancer is one of the most severe diseases nowadays. Thus, tumor detection in a non-invasive and accurate manner is a challenging subject. Among these tumors, liver cancer is one of the most dangerous, being very common. Hepatocellular Carcinoma (HCC) is the most frequent malignant liver tumor. The golden standard for diagnosing HCC is mainly the biopsy, however invasive and risky, leading to infections, respectively to the spreading of the tumor through the body. We conceive computerized techniques for abdominal tumor recognition within medical images. Formerly, traditional, texture-based methods were involved for this purpose. Both classical texture analysis methods, as well as advanced, original texture analysis techniques, based on superior order statistics, were involved. The superior order Gray Level Cooccurrence Matrix (GLCM), as well as the Textural Microstructure Cooccurrence Matrices (TMCM) were employed and assessed. Recently, deep learning techniques based on Convolutional Neural Networks (CNN), their fusions with the conventional techniques, as well as their combinations among themselves, were assessed in the approached field. We present the most relevant aspects of this study in the current paper.

Real-Time Obstacle Detection using a Pillar-based Representation and a Parallel Architecture on the GPU from LiDAR Measurements

Authors: Sergiu Nedevschi, Mircea Paul Mureșan, Radu Gabriel Dănescu

Software Program for the Evaluation of Human Exposure to Electric and Magnetic Fields

Authors: Călin Munteanu, Vasile Topa

The evaluation of human exposure to electric and magnetic fields represents a subject of great scientific and public interest due to the biological effects of electromagnetic fields (EMFs) on the human body and the risks caused by them to living organisms. In this context, this article proposes a software program designed by the authors for the evaluation of human exposure to electric and magnetic fields at low frequencies (EMF software program), an application that can also be accessed from a mobile phone. The analytical model on which the EMF program is based is synthetically presented, and the application is then described. The first example implemented in the EMF program is taken from the existing literature on this subject, thus confirming the correctness and calculation precision of the program. Next, a case study is proposed for an overhead transmission line of 400 KV from the Cluj-Napoca area, Romania, for which the electric and magnetic fields are first measured experimentally and then using the EMF program. The validation of the EMF software program is performed by comparing the obtained results with those measured experimentally and with those obtained with a commercial software program.

Implementation of an Accurate Measurement Method for the Spatial Distribution of the Electromagnetic Field in a WPT System

Authors: Vasile Topa

In an inductive wireless power transfer (WPT) system, the transferred power depends on the coupling between the transmitter coil and the receiver coil and, consequently, on their spatial positioning with respect to each other. Since this positioning needs to be as flexible as possible, one can design various construction shapes for the coils, in terms of both winding geometry and core geometry. The system efficiency for different positionings and configurations can be assessed if the structure and distribution of the electromagnetic field (EMF) in the space close to the system are determined. In this context, this paper proposes an investigation method based on measuring the intensity of the electromagnetic field generated by the emitting coil in the WPT space in terms of both modulus and direction. The measurements needed to be performed at a sufficient number of points, so that the entire field was determined. The EMF value was estimated by linear spatial interpolation between the measured points. Moreover, this paper presents the method and the results of investigations carried out only in the presence of the emitting coil. Calibration or correction solutions were also proposed to obtain sufficient accuracy, and the viability of the method was therefore demonstrated.

The Optimization of the Interior Permanent Magnetic Motor Case Study

Authors: Vasile Topa

This paper presents two optimization methods of an interior permanent magnetic motor (IPM). The first method is based on the non-dominated sorting genetic algorithm II (NSGA-II), while the second one is based on the NSGA-III algorithm. The focus is on the reduction of the cogging torque (CT), which acts upon the rotor when the IPM is not powered, and the current that flows through the three windings is 0. The influence of the CT over the motor operation is a negative one, inducing vibrations that are not desired. The limitation of these influences is the desired way for improvement. The optimizations applied entailed the change of the geometrical configuration of the stator whilst maintaining the exterior dimensions unaltered. Through this simple but robust approach, the performance of the IPM proposed was improved.

Spike sorting using Superlets: Evaluation of a novel feature space for the discrimination of neuronal spikes

Authors: Eugen Richard Ardelean

Because of the intricacy of neural data, spike sorting is a challenging problem in neuroscience. Overlapping clusters are one of the hardest issues to solve, along with the similarity of spike shapes, excessive noise, and unbalanced clusters. Here, we focus on analyzing the efficacy and integrating the Superlets Transform as a feature extraction method. Our results indicate that Superlets achieve not only super-resolution in time and frequency, but also provide the conditions for smooth clustering performance. The Superlet Transform was analyzed in conjunction with Principal Component Analysis, Singular Value Decomposition, and Isomap Embedding as dimensionality reduction methods. The Superlet-extracted features were classified using Neural Networks in order to assess their relevance.

Local Field Potential Microstate Analysis

Authors: Eugen Richard Ardelean

Correct depth of anesthesia (DOA) assessment is a serious and widespread medical problem and an active scientific research topic. Anesthesia is known to alter the dynamics of neural networks within the brain that ultimately results in the impairment of pain perception. Nevertheless, linking the composition and dosage of various anesthetics, to a target level of anesthesia is not trivial. In this paper we explore the use of microstates as a viable tool to discriminate between anesthesia levels in intracranial recordings from mouse visual cortex. We show such symbolic analysis is able to capture DOA specific information in local field potentials (LFP). Microstates are characterized by the appearance of set of prototypical maps of activations over recording sites (electrodes) that define stereotypical patterns of activation in the recorded neural networks. Although microstates have been defined and characterized in electroencephalogram (EEG) data, our study shows for the first time that microstates can be effective when considering LFPs as well. We performed statistical analysis of average duration, time coverage, and occurrence of the microstates in order to differentiate between different DOA levels. By increasing the number of microstates, the analysis is more insightful and it is easier to discriminate between DOA levels.

Geographies of Capitalism in Romanian Postcommunist Literature and Film: Villages, Cities, Markets

Authors: Adriana Stan

This article analyses the ways in which the rural – understood in its double nature, of traditional mythological space, and of degraded periphery of the urban – was both instrumentalized by the cultural and political discourses within Romanian postcommunism, and revealed as a symptomatic field in which the harsh economic realities of the post-1989 implementation of neoliberal capitalism in Eastern Europe became evident. The first part of the article describes how misplaced public policies during 1980s’ communism favoured the underdevelopment of the rural. Later, the process of transition encouraged both external migration, which depopulated Romanian villages, and the internal migration of the urban middle-class to an economically more secure rural area. The second part of the article analyses – from a materialist viewpoint – corresponding topics developed in Romanian contemporary literature and films (authored by Adrian Șchiop, Lavinia Braniște, Bogdan Mirică, and Radu Muntean). These works envision the rural/periurban as an asymmetrical geography, where violence toward the vulnerable is enacted, where the loss of community values leaves black markets flourish, and the capitalist dismantling of social ties becomes striking.

Predicting Visual Acuity in Patients Treated for AMD

Authors: Adrian Petru Groza

The leading diagnostic tool in modern ophthalmology, Optical Coherence Tomography (OCT), is not yet able to establish the evolution of retinal diseases. Our task is to forecast the progression of retinal diseases by means of machine learning technologies. The aim is to help the ophthalmologist to determine when early treatment is needed in order to prevent severe vision impairment or even blindness. The acquired data are made up of sequences of visits from multiple patients with age-related macular degeneration (AMD), which, if not treated at the appropriate time, may result in irreversible blindness. The dataset contains 94 patients with AMD and there are 161 eyes included with more than one medical examination. We used various techniques from machine learning (linear regression, gradient boosting, random forest and extremely randomised trees, bidirectional recurrent neural network, LSTM network, GRU network) to handle technical challenges such as how to learn from small-sized time series, how to handle different time intervals between visits, and how to learn from different numbers of visits for each patient (1-5 visits). For predicting the visual acuity, we performed several experiments with different features. First, by considering only previous measured visual acuity, the best accuracy of 0.96 was obtained based on a linear regression. Second, by considering numerical OCT features such as previous thickness and volume values in all retinal zones, the LSTM network reached the highest score (R2=0.99). Third, by considering the fundus scan images represented as embeddings obtained from the convolutional autoencoder, the accuracy was increased for all algorithms. The best forecasting results for visual acuity depend on the number of visits and features used for predictions, i.e., 0.99 for LSTM based on three visits (monthly resampled series) based on numerical OCT values, fundus images, and previous visual acuities.

An Enhanced Evaporation Rate Water-Cycle Algorithm for Global Optimization

Authors: Adrian Pop

Water-cycle algorithm based on evaporation rate (ErWCA) is a powerful enhanced version of the water-cycle algorithm (WCA) metaheuristics algorithm. ErWCA, like other algorithms, may still fall in the sub-optimal region and have a slow convergence, especially in high-dimensional tasks problems. This paper suggests an enhanced ErWCA (EErWCA) version, which embeds local escaping operator (LEO) as an internal operator in the updating process. ErWCA also uses a control-randomization operator. To verify this version, a comparison between EErWCA and other algorithms, namely, classical ErWCA, water cycle algorithm (WCA), butterfly optimization algorithm (BOA), bird swarm algorithm (BSA), crow search algorithm (CSA), grasshopper optimization algorithm (GOA), Harris Hawks Optimization (HHO), whale optimization algorithm (WOA), dandelion optimizer (DO) and fire hawks optimization (FHO) using IEEE CEC 2017, was performed. The experimental and analytical results show the adequate performance of the proposed algorithm.

A Modular, Extensible, and Modelica-Standard-Compliant OpenModelica Compiler Framework in Julia Supporting Structural Variability

Authors: Adrian Pop

Nowadays, industrial products are getting increasingly complex, and time-to-market is significantly shorter. Modeling and simulation tools for cyber-physical systems need to keep up with the increased complexity. This paper presents OpenModelica.jl, a modular and extensible Modelica compiler framework in Julia targeting ModelingToolkit.jl and supporting Variable Structured Systems. We extended the Modelica language with three new operators to support continuous-time mode-switching and reconfiguration via recompilation at runtime. Therefore, our compiler supports the Modelica language and variable structure systems via the aforementioned extensions. To our knowledge, there are no other Modelica tools available that support both standard Modelica and variable structure systems. We evaluated our framework using a standardized benchmark suite, in terms of simulation, compilation and recompilation performance. The results concerning compilation and simulation time performance were compared with the results of running the existing OpenModelica compiler with the same set of models. A custom benchmark was devised to estimate the cost in terms of recompilation when simulating variable structure systems. The performance experiments showed that OpenModelica.jl is currently about four times slower in terms of compilation time when compiling a transmission line model with tens of thousands of equations and variables. The difference in simulation performance between the two compilers was negligable. Furthermore, the impact of recompilation during the simulation was usually small compared with the simulation time for long simulations. The results are promising for a prototype, and we outline approaches to further improve both compilation and simulation performance as future research.

Property Probes: Source Code Based Exploration of Program Analysis Results

Authors: Adrian Pop

We present property probes, a mechanism for helping a developer interactively explore partial program analysis results in terms of the source program, and as the program is edited. A node locator data structure is introduced that maps between source code spans and program representation nodes, and that helps identify probed nodes in a robust way, after modifications to the source code. We have developed a client-server based tool supporting property probes, and argue that it is very helpful in debugging and understanding program analyses. We have evaluated our tool on several languages and analyses, including a full Java compiler and a tool for intraprocedural dataflow analysis. Our performance results show that the probe overhead is negligible even when analyzing large projects.

FlexLip: A Controllable Text-to-Lip System

Authors: Adriana Stan

The task of converting text input into video content is becoming an important topic for synthetic media generation. Several methods have been proposed with some of them reaching close-to-natural performances in constrained tasks. In this paper, we tackle a subissue of the text-to-video generation problem, by converting the text into lip landmarks. However, we do this using a modular, controllable system architecture and evaluate each of its individual components. Our system, entitled FlexLip, is split into two separate modules: text-to-speech and speech-to-lip, both having underlying controllable deep neural network architectures. This modularity enables the easy replacement of each of its components, while also ensuring the fast adaptation to new speaker identities by disentangling or projecting the input features. We show that by using as little as 20 min of data for the audio generation component, and as little as 5 min for the speech-to-lip component, the objective measures of the generated lip landmarks are comparable with those obtained when using a larger set of training samples. We also introduce a series of objective evaluation measures over the complete flow of our system by taking into consideration several aspects of the data and system configuration. These aspects pertain to the quality and amount of training data, the use of pretrained models, and the data contained therein, as well as the identity of the target speaker; with regard to the latter, we show that we can perform zero-shot lip adaptation to an unseen identity by simply updating the shape of the lips in our model.

Residual Information in Deep Speaker Embedding Architectures

Authors: Adriana Stan

Speaker embeddings represent a means to extract representative vectorial representations from a speech signal such that the representation pertains to the speaker identity alone. The embeddings are commonly used to classify and discriminate between different speakers. However, there is no objective measure to evaluate the ability of a speaker embedding to disentangle the speaker identity from the other speech characteristics. This means that the embeddings are far from ideal, highly dependent on the training corpus and still include a degree of residual information pertaining to factors such as linguistic content, recording conditions or speaking style of the utterance. This paper introduces an analysis over six sets of speaker embeddings extracted with some of the most recent and high-performing deep neural network (DNN) architectures, and in particular, the degree to which they are able to truly disentangle the speaker identity from the speech signal. To correctly evaluate the architectures, a large multi-speaker parallel speech dataset is used. The dataset includes 46 speakers uttering the same set of prompts, recorded in either a professional studio or their home environments. The analysis looks into the intra- and inter-speaker similarity measures computed over the different embedding sets, as well as if simple classification and regression methods are able to extract several residual information factors from the speaker embeddings. The results show that the discriminative power of the analyzed embeddings is very high, yet across all the analyzed architectures, residual information is still present in the representations in the form of a high correlation to the recording conditions, linguistic contents and utterance duration. However, we show that this correlation, although not ideal, could still be useful in downstream tasks. The low-dimensional projections of the speaker embeddings show similar behavior patterns across the embedding sets with respect to intra-speaker data clustering and utterance outlier detection.

The ZevoMOS entry to VoiceMOS Challenge 2022

Authors: Adriana Stan

This paper introduces the ZevoMOS entry to the main track of the VoiceMOS Challenge 2022. The ZevoMOS submission is based on a two-step finetuning of pretrained self-supervised learning (SSL) speech models. The first step uses a task of classifying natural versus synthetic speech, while the second step's task is to predict the MOS scores associated with each training sample. The results of the finetuning process are then combined with the confidence scores extracted from an automatic speech recognition model, as well as the raw embeddings of the training samples obtained from a wav2vec SSL speech model. The team id assigned to the ZevoMOS system within the VoiceMOS Challenge is T01. The submission was placed on the 14th place with respect to the system-level SRCC, and on the 9th place with respect to the utterance-level MSE. The paper also introduces additional evaluations of the intermediate results.

Predicting the Geoeffectiveness of CMEs Using Machine Learning

Authors: Anca Nicoleta Mărginean

Abstract Coronal mass ejections (CMEs) are the most geoeffective space weather phenomena, being associated with large geomagnetic storms, and having the potential to cause disturbances to telecommunications, satellite network disruptions, and power grid damage and failures. Thus, considering these storms’ potential effects on human activities, accurate forecasts of the geoeffectiveness of CMEs are paramount. This work focuses on experimenting with different machine-learning methods trained on white-light coronagraph data sets of close-to-Sun CMEs, to estimate whether such a newly erupting ejection has the potential to induce geomagnetic activity. We developed binary classification models using logistic regression, k-nearest neighbors, support vector machines, feed-forward artificial neural networks, and ensemble models. At this time, we limited our forecast to exclusively use solar onset parameters, to ensure extended warning times. We discuss the main challenges of this task, namely, the extreme imbalance between the number of geoeffective and ineffective events in our data set, along with their numerous similarities and the limited number of available variables. We show that even in such conditions adequate hit rates can be achieved with these models.

Virtual Power Plant Optimization in Smart Grids: A Narrative Review

Authors: Tudor Cioara, Anghel Ionuț

Virtual power plants (VPPs) are promising solutions to address the decarbonization and energy efficiency goals in the smart energy grid. They assume the coordination of local energy resources such as energy generation, storage, and consumption. They are used to tackle problems brought by the stochastic nature of renewable energy, lack of energy storage devices, or insufficient local energy flexibility on the demand side. VPP modeling, management, and optimization are open to research problems that should consider, on one side, the local constraints in the operation of the energy resources and power flows and the energy grid’s sustainability objectives on the other side. There are multiple goals to create a VPP, such as to deliver energy services on a market or to the grid operator, to operate a microgrid in autonomy decoupled from the main grid, or to sustain local energy communities. In this paper, we present the results of a narrative review carried out on the domain of VPP optimization for the local energy grid integration. We have defined a search strategy that considers highly rated international databases (i.e., Elsevier, IEEE, and MDPI) in a six-year timeframe and applied objective inclusion/exclusion criteria for selecting articles and publications for the review; 95 articles have been analyzed and classified according to their objectives and solutions proposed for optimizing VPP integration in smart grids. The results of the study show that VPP concepts and applications are well addressed in the research literature, however, there is still work to be done on: engaging prosumers and citizens in such a virtual organization, developing heuristics to consider a wider range of local and global constraints and non-energy vectors, and to decentralize and make transparent the services delivery and financial settlement towards community members. This study can help researchers to understand the current directions for VPP integration in smart grids. As a next step we plan to further analyze the open research directions related to this problem and target the development of innovative solutions to allow the integration of multi-energy assets and management of cross energy sector services in energy communities.

Review of bio-inspired optimization applications in renewable-powered smart grids: Emerging population-based metaheuristics

Authors: Tudor Cioara, Anghel Ionuț

The management of renewable-powered smart grids deals with nonlinear optimization problems featuring a variety of linear or nonlinear constraints, discrete or continuous optimization variables, involving high dimensionality of the solution space, and strict time requirements to identify the optimal or near-optimal solution. One promising approach for addressing such optimization problems is to apply bio-inspired population-based optimization algorithms, many such metaheuristics emerging lately. In this paper, we have identified the metaheuristics with the highest impact published recently and reviewed their applications in the management of renewable-powered smart energy grids using the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) methodology and the Web of Science Core Collection as the reference database. Four main smart grid application domains we been analyzed: (i) energy prediction models’ optimization to reduce uncertainty (ii) energy resources coordination to handle the stochastic nature of renewables, (iii) demand response using controllable loads and flexibility while considering the consumers’ needs and constraints and (iv) optimization of grid energy efficiency and costs. The results showed the advantages of such metaheuristics for decentralized optimization problems with low computational time and resource overhead. At the same time, several issues need to be addressed to increase their adoption in the smart grid management scenarios: the lack of standard testing methodologies and benchmarks, efficient management of exploration and exploitation of the optimization search space, guidelines for metaheuristics application with clear links to the type of optimization problems, etc.

Blockchain-Based Distributed Federated Learning in Smart Grid

Authors: Tudor Cioara, Anghel Ionuț

The participation of prosumers in demand-response programs is essential for the success of demand-side management in renewable-powered energy grids. Unfortunately, the engagement is still low due to concerns related to the privacy of their energy data used in the prediction processes. In this paper, we propose a blockchain-based distributed federated learning (FL) technique for energy-demand prediction that combines FL with blockchain to provide data privacy and trust features for energy prosumers. The privacy-sensitive energy data are stored locally at edge prosumer nodes without revealing it to third parties, with only the learned local model weights being shared using a blockchain network. The global federated model is not centralized but distributed and replicated over the blockchain overlay, ensuring the model immutability and provenance of parameter updates. We had proposed smart contracts to deal with the integration of local machine-learning prediction models with the blockchain, defining functions for the model parameters’ scaling and reduction of blockchain overhead. The centralized, local-edge, and blockchain-integrated models are comparatively evaluated for prediction of energy demand 24 h ahead using a multi-layer perceptron model and the monitored energy data of several prosumers. The results show only a slight decrease in prediction accuracy in the case of blockchain-based distributed FL with reliable data privacy support compared with the centralized learning solution.

Identifying and Monitoring the Daily Routine of Seniors Living at Home

Authors: Tudor Cioara, Anghel Ionuț

As the population in the Western world is rapidly aging, the remote monitoring solutions integrated into the living environment of seniors have the potential to reduce the care burden helping them to self-manage problems associated with old age. The daily routine is considered a useful tool for addressing age-related problems having additional benefits for seniors like reduced stress and anxiety, increased feeling of safety and security. In this paper, we propose a solution for identifying the daily routines of seniors using the monitored activities of daily living and for inferring deviations from the routines that may require caregivers' interventions. A Markov model-based method is defined to identify the daily routines, while entropy rate and cosine functions are used to measure and assess the similarity between the daily monitored activities in a day and the inferred routine. A distributed monitoring system was developed that uses Beacons and trilateration techniques for monitoring the activities of older adults. The results are promising, the proposed techniques can identify the daily routines with confidence concerning the activity duration of 0.98 and the sequence of activities in the interval of [0.0794, 0.0829]. Regarding deviation identification, our method obtains 0.88 as the best sensitivity value with an average precision of 0.95.

A Review of Blockchain Technology Applications in Ambient Assisted Living

Authors: Tudor Cioara, Anghel Ionuț

The adoption of remote assisted care was accelerated by the COVID-19 pandemic. This type of system acquires data from various sensors, runs analytics to understand people’s activities, behavior, and living problems, and disseminates information with healthcare stakeholders to support timely follow-up and intervention. Blockchain technology may offer good technical solutions for tackling Internet of Things monitoring, data management, interventions, and privacy concerns in ambient assisted living applications. Even though the integration of blockchain technology with assisted care is still at the beginning, it has the potential to change the health and care processes through a secure transfer of patient data, better integration of care services, or by increasing coordination and awareness across the continuum of care. The motivation of this paper is to systematically review and organize these elements according to the main problems addressed. To the best of our knowledge, there are no studies conducted that address the solutions for integrating blockchain technology with ambient assisted living systems. To conduct the review, we have followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology with clear criteria for including and excluding papers, allowing the reader to effortlessly gain insights into the current state-of-the-art research in the field. The results highlight the advantages and open issues that would require increased attention from the research community in the coming years. As for directions for further research, we have identified data sharing and integration of care paths with blockchain, storage, and transactional costs, personalization of data disclosure paths, interoperability with legacy care systems, legal issues, and digital rights management.

Hybrid Deep Neural Network Model for Multi-Step Energy Prediction of Prosumers

Authors: Tudor Cioara, Anghel Ionuț

The transition toward decentralized renewable energy transforms the energy grid, with the prosumers playing an active role in the local energy management. Accurate day-ahead prediction of their energy demand is a prerequisite to ensure the stability and efficiency of the power grid by balancing the energy demand with the production, while incorporating renewable energy as much as possible. In this paper, we propose a 24-steps-ahead energy prediction model that integrates clustering and multilayer perceptron classification models used to detect the classes of energy profiles and multilayer perceptron regression models used to fine-tune the energy prediction, considering the energy data streams. We introduce new features derived from the raw energy data collected from prosumers, such as the profile peaks and valleys, concerning the energy baseline and describe a software infrastructure for integrating the real-time energy data streams with the hybrid deep learning models training and prediction. The evaluation tests consider energy datasets that are closer to the real-time energy data streams from prosumers. The results show that, even on energy data streams, the model offers a good prediction accuracy for small- and medium-scale prosumers.

Deep Learning, Mining, and Collaborative Clustering to Identify Flexible Daily Activities Patterns

Authors: Tudor Cioara, Anghel Ionuț, Alexandru Rancea

The monitoring of the daily life activities routine is beneficial, especially in old age. It can provide relevant information on the person's health state and wellbeing and can help identify deviations that signal care deterioration or incidents that require intervention. Existing approaches consider the daily routine as a rather strict sequence of activities which is not usually the case. In this paper, we propose a solution to identify flexible daily routines of older adults considering variations related to the order of activities and activities timespan. It combines the Gap-BIDE algorithm with a collaborative clustering technique. The Gap-BIDE algorithm is used to identify the most common patterns of behavior considering the elements of variations in activities sequence and the period of the day (i.e., night, morning, afternoon, and evening) for increased pattern mining flexibility. K-means and Hierarchical Clustering Agglomerative algorithms are collaboratively used to address the time-related elements of variability in daily routine like activities timespan vectors. A prototype was developed to monitor and detect the daily living activities based on smartwatch data using a deep learning architecture and the InceptionTime model, for which the highest accuracy was obtained. The results obtained are showing that the proposed solution can successfully identify the routines considering the aspects of flexibility such as activity sequences, optional and compulsory activities, timespan, and start and end time. The best results were obtained for the collaborative clustering solution that considers flexibility aspects in routine identification, providing coverage of monitored data of 89.63%.

Identification of Daily Living Recurrent Behavioral Patterns Using Genetic Algorithms for Elderly Care

Authors: Tudor Cioara, Anghel Ionuț

A person’s routine is a sequence of activities of daily living patterns recurrently performed. Sticking daily routines is a great tool to support the care of persons with dementia, and older adults in general, who are living in their homes, and also being useful for caregivers. As state-of-the-art tools based on self-reporting are subjective and rely on a person’s memory, new tools are needed for objectively detecting such routines from the monitored data coming from wearables or smart home sensors. In this paper, we propose a solution for detecting the daily routines of a person by extracting the sequences of recurrent activities and their duration from the monitored data. A genetic algorithm is defined to extract activity patterns featuring small differences that relate to the day-to-day contextual variations that occur in a person’s daily routine. The quality of the solutions is evaluated with a probabilistic-based fitness function, while a tournament-based strategy is employed for the dynamic selection of mutation and crossover operators applied for generating the offspring. The time variability of activities of daily living is addressed using the dispersion of the values of duration of that activity around the average value. The results are showing an accuracy above 80% in detecting the routines, while the optimal values of population size and the number of generations for fitness function evolution and convergence are determined using multiple linear regression analysis.

Emerging Sensors Techniques and Technologies for Intelligent Environments

Authors: Tudor Cioara, Anghel Ionuț

The trending techniques for managing indoor and outdoor intelligent environments rely heavily on data acquisition through a diversity of heterogeneous Internet of Things (IoT) devices and sensors [...].

Exploration-Based Search for an Unknown Number of Targets using a UAV

Authors: Lucian Bușoniu, Zsofia Lendek

We consider a scenario in which a UAV must locate an unknown number of targets at unknown locations in a 2D environment. A random finite set formulation with a particle filter is used to estimate the target locations from noisy measurements that may miss targets. A novel planning algorithm selects a next UAV state that maximizes an objective function consisting of two components: target refinement and an exploration. Found targets are saved and then disregarded from measurements to focus on refining poorly seen targets. The desired next state is used as a reference point for a nonlinear tracking controller for the robot. Simulation results show that the method works better than lawnmower and mutual-information baselines.

Online learning control for path-aware global optimization with nonlinear mobile robots

Authors: Lucian Bușoniu

Consider a robot with nonlinear dynamics that must quickly find a global optimum of an objective function defined over its operating area, e.g., a chemical concentration, physical measurement, quantity of material etc. The function is initially unknown and must be learned online from samples acquired in a single trajectory. Applying classical optimization methods in this scenario would be highly suboptimal, since they would place the next sample arbitrarily far, without taking into account robot motion constraints, and would not revise the path based on new information accumulated along it. To address these limitations, a novel algorithm called Path-Aware Optimistic Optimization (OOPA) is proposed. The decision of which robot action to apply is formulated as an optimal control problem in which the rewards are refinements of the upper bound on the objective, weighted by bound and objective values to focus the search around optima. OOPA is evaluated in extensive simulations where it is compared to path-unaware optimization baselines, and in a real experiment in which a ROBOTIS TurtleBot3 successfully searches for the lowest grayscale location on a 2D surface.

A Simulator and First Reinforcement Learning Results for Underwater Mapping

Authors: Lucian Bușoniu

Underwater mapping with mobile robots has a wide range of applications, and good models are lacking for key parts of the problem, such as sensor behavior. The specific focus here is the huge environmental problem of underwater litter, in the context of the Horizon 2020 SeaClear project, where a team of robots is being developed to map and collect such litter. No reinforcement-learning solution to underwater mapping has been proposed thus far, even though the framework is well suited for robot control in unknown settings. As a key contribution, this paper therefore makes a first attempt to apply deep reinforcement learning (DRL) to this problem by exploiting two state-of-the-art algorithms and making a number of mapping-specific improvements. Since DRL often requires millions of samples to work, a fast simulator is required, and another key contribution is to develop such a simulator from scratch for mapping seafloor objects with an underwater vehicle possessing a sonar-like sensor. Extensive numerical experiments on a range of algorithm variants show that the best DRL method collects litter significantly faster than a baseline lawn mower trajectory.

Underwater Robot Pose Estimation Using Acoustic Methods and Intermittent Position Measurements at the Surface

Authors: Lucian Bușoniu, Levente Tamás

Global positioning systems can provide sufficient positioning accuracy for large scale robotic tasks in open environments. However, in underwater environments, these systems cannot be directly used, and measuring the position of underwater robots becomes more difficult. In this paper we first evaluate the performance of existing pose estimation techniques for an underwater robot equipped with commonly used sensors for underwater control and pose estimation, in a simulated environment. In our case these sensors are inertial measurement units, Doppler velocity log sensors, and ultra-short baseline sensors. Secondly, for situations in which underwater estimation suffers from drift, we investigate the benefit of intermittently correcting the position using a high-precision surface-based sensor, such as regular GPS or an assisting unmanned aerial vehicle that tracks the underwater robot from above using a camera.

A simple path-aware optimization method for mobile robots*

Authors: Lucian Bușoniu

We present an approach for a mobile robot to seek the global maximum of an initially unknown function defined over its operating space. The method exploits a Lipschitz assumption to defne an upper bound on the function from previously seen samples, and optimistically moves towards the largest upper-bound point. This point is iteratively changed whenever new samples make it clear that it is suboptimal. In simulations, the method finds the global maxima with much less computation than an existing, much more involved technique, while keeping performance acceptable. Real-robot experiments confirm the effectiveness of the approach.

Stability analysis of optimal control problems with time-dependent costs

Authors: Lucian Bușoniu

We present stability conditions for deterministic time-varying nonlinear discrete-time systems whose inputs aim to minimize an infinite-horizon time-dependent cost. Global asymptotic and exponential stability properties for general attractors are established. This work covers and generalizes the related results on discounted optimal control problems to more general systems and cost functions.

Enhanced Perception for Autonomous Driving Using Semantic and Geometric Data Fusion

Authors: Sergiu Nedevschi, Andra Petrovai

Environment perception remains one of the key tasks in autonomous driving for which solutions have yet to reach maturity. Multi-modal approaches benefit from the complementary physical properties specific to each sensor technology used, boosting overall performance. The added complexity brought on by data fusion processes is not trivial to solve, with design decisions heavily influencing the balance between quality and latency of the results. In this paper we present our novel real-time, 360∘ enhanced perception component based on low-level fusion between geometry provided by the LiDAR-based 3D point clouds and semantic scene information obtained from multiple RGB cameras, of multiple types. This multi-modal, multi-sensor scheme enables better range coverage, improved detection and classification quality with increased robustness. Semantic, instance and panoptic segmentations of 2D data are computed using efficient deep-learning-based algorithms, while 3D point clouds are segmented using a fast, traditional voxel-based solution. Finally, the fusion obtained through point-to-image projection yields a semantically enhanced 3D point cloud that allows enhanced perception through 3D detection refinement and 3D object classification. The planning and control systems of the vehicle receives the individual sensors' perception together with the enhanced one, as well as the semantically enhanced 3D points. The developed perception solutions are successfully integrated onto an autonomous vehicle software stack, as part of the UP-Drive project.

Semantic Cameras for 360-Degree Environment Perception in Automated Urban Driving

Authors: Sergiu Nedevschi, Andra Petrovai

The European UP-Drive project addresses transportation-related challenges by providing key contributions that enable fully automated vehicle navigation and parking in complex urban areas, which results in a safer, inclusive, affordable and environmentally friendly transportation system. For this purpose, the project consortium developed a prototype electrical vehicle equipped with cameras and LiDARs sensors that is capable to autonomously drive around the city and find available parking spots. In UP-Drive, we created an accurate, robust and redundant multi-modal environment perception system that provides 360° coverage around the vehicle. This paper summarizes the work of the project related to the surround view semantic perception using fisheye and narrow field-of-view semantic virtual cameras. Deep learning-based semantic, instance and panoptic segmentation networks, which satisfy requirements in accuracy and efficiency have been developed and integrated into the final prototype. The UP-Drive automated vehicle has been successfully demonstrated in urban areas after extensive experiments and numerous field tests.

Fast Panoptic Segmentation with Soft Attention Embeddings

Authors: Sergiu Nedevschi, Andra Petrovai

Panoptic segmentation provides a rich 2D environment representation by unifying semantic and instance segmentation. Most current state-of-the-art panoptic segmentation methods are built upon two-stage detectors and are not suitable for real-time applications, such as automated driving, due to their high computational complexity. In this work, we introduce a novel, fast and accurate single-stage panoptic segmentation network that employs a shared feature extraction backbone and three network heads for object detection, semantic segmentation, instance-level attention masks. Guided by object detections, our new panoptic segmentation head learns instance specific soft attention masks based on spatial embeddings. The semantic masks for stuff classes and soft instance masks for things classes are pixel-wise coherent and can be easily integrated in a panoptic output. The training and inference pipelines are simplified and no post-processing of the panoptic output is necessary. Benefiting from fast inference speed, the network can be deployed in automated vehicles or robotic applications. We perform extensive experiments on COCO and Cityscapes datasets and obtain competitive results in both accuracy and time. On the Cityscapes dataset we achieve 59.7 panoptic quality with an inference speed of more than 10 FPS on high resolution 1024 × 2048 images.

Time-Space Transformers for Video Panoptic Segmentation

Authors: Sergiu Nedevschi, Andra Petrovai

We propose a novel solution for the task of video panoptic segmentation, that simultaneously predicts pixel-level semantic and instance segmentation and generates clip-level instance tracks. Our network, named VPS-Transformer, with a hybrid architecture based on the state-of-the-art panoptic segmentation network Panoptic-DeepLab, combines a convolutional architecture for single-frame panoptic segmentation and a novel video module based on an instantiation of the pure Transformer block. The Transformer, equipped with attention mechanisms, models spatio-temporal relations between backbone output features of current and past frames for more accurate and consistent panoptic estimates. As the pure Transformer block introduces large computation overhead when processing high resolution images, we propose a few design changes for a more efficient compute. We study how to aggregate information more effectively over the space-time volume and we compare several variants of the Transformer block with different attention schemes. Extensive experiments on the Cityscapes-VPS dataset demonstrate that our best model improves the temporal consistency and video panoptic quality by a margin of 2.2%, with little extra computation.

MonoDVPS: A Self-Supervised Monocular Depth Estimation Approach to Depth-aware Video Panoptic Segmentation

Authors: Sergiu Nedevschi, Andra Petrovai

Depth-aware video panoptic segmentation tackles the inverse projection problem of restoring panoptic 3D point clouds from video sequences, where the 3D points are augmented with semantic classes and temporally consistent instance identifiers. We propose a novel solution with a multi-task network that performs monocular depth estimation and video panoptic segmentation. Since acquiring ground truth labels for both depth and image segmentation has a relatively large cost, we leverage the power of unlabeled video sequences with self-supervised monocular depth estimation and semi-supervised learning from pseudo-labels for video panoptic segmentation. To further improve the depth prediction, we introduce panoptic-guided depth losses and a novel panoptic masking scheme for moving objects to avoid corrupting the training signal. Extensive experiments on the Cityscapes-DVPS and SemKITTI-DVPS datasets demonstrate that our model with the proposed improvements achieves competitive results and fast inference speed.

A Novel Multi-Scale Method for Thermo-Mechanical Simulation of Power Integrated Circuits

Authors: Vasile Topa

During development of power Integrated Circuits (IC), several iterations between the design and test/ measurement steps are performed. Computer-aided engineering significantly shortens the product development process because the numerical simulations can identify and remediate most deficiencies during the design stage. The recent IC manufacturing technologies lead to ca.$10^{4}$-order scale separation between transistor cell details and the device active area, resulting in very complex IC models. For the IC complexity to be overcome, advanced multi-scale analysis methods are required to perform accurate simulations in a decent time (order of hours). This paper proposes an advanced and enhanced multi-scale simulation method for the thermo-mechanical analysis of power ICs. The computational IC structure is automatically generated from a Cadence layout and partitioned into far-field and homogenized regions - the macro-model. Detailed localized micro-scale sub-models are assigned to limited portions of the homogenized region. The two-way simulated data transfer between the homogenized macro-model and the micro sub-models is one multi-scale approach novelty proposed in this paper. The method is validated on a real test chip structure presented in literature. The proposed multi-scale approach in conjunction with the two-way macro-micro data transfer lead to similar accuracy in the prediction of defect location, yet with significant simulation time - and computational resource reduction (CPU time and RAM usage reduced by almost 80% and 60% respectively) compared to the method used as reference.

The influence of the patch antennas emissions on the human head

Authors: Vasile Topa, Călin Munteanu

Abstract The influences of the antenna’s emissions on the human body are to be investigated because in our days, as a simple example, the mobile phone is permanently in use, on day or night. The long-time human exposure at the electromagnetic field radiations is obvious and there exist the possibility that these can affect the human health. The influences of the circular patch antenna’s emissions on the human head are analyzed at high frequency in this paper. First the patch antenna is modelled and simulated to extract its specific parameters and to represent the emitted electric and magnetic field around it. Next the human head is considered near to the patch antenna in position that imitates the mobile phone at ear and the electric field, the magnetic field, and the specific absorption rate (SAR) are represented to determine the antenna’s emissions at its operating frequency. So, the main aim of the paper is the high frequency analyzes of the circular patch antennas to investigate the influence of their emissions on the human head. The results obtained are compared and reported to the standards and legislation in force on human occupational exposure, to verify that the limits are met.

Test Structure Design for Defect Detection during Active Thermal Cycling

Authors: Vasile Topa

Integrated power ICs acting as smart power switches for automotive or industrial applications are often subjected to active thermal cycling. Consequently, they undergo significant self-heating and are prone to various failure mechanisms related to the electro-thermo-mechanical phenomena that take place in the device metallization. In this article a test structure consisting of a lateral DMOS transistor equipped with several integrated sensors is proposed for metallization fatigue assessment. The design of the test structure is presented in detail, alongside with design considerations drawn from the literature and from simulation results. The testing procedure is then described, and experimental results are discussed. The experimental data provided by the integrated sensors correlated with the electro-thermal simulation results indicate the emergence of a failure mechanism and this is later confirmed by failure analysis. Conclusions are further drawn regarding the feasibility of using the proposed integrated sensors for monitoring defects in power ICs.

Emerging policies for the regulation of AI

Authors: Adrian Petru Groza, Camelia Lemnaru

The report analyzes emerging policies for the regulation of artificial intelligence (AI) and their interaction with existing regulations in related fields. The report has two parts. The first part examines the European Commission’s regulatory framework for artificial intelligence, its integration into emerging policies for the development of ethical AI, and its interaction with regulations in other sectors. The second part presents a possible institutional architecture for the regulation of AI in Romania. This vision involves a competent national authority for monitoring and regulating ethical AI, as well as a decentralized model in which conformity assessment bodies (e.g. audit centers, whether private or public) have responsibilities for monitoring, verification, and certification in different subdomains and technologies of artificial intelligence.

The experience of the limit or the failure of the human: A philosophical anthropology against the backdrop of the sacred-profane dialectic

Authors: Adrian Petru Groza

The text aims to analyze modes of subjectivation in the context of the meta-experience of the sacred-profane dialectic, particularly in the context of the full concealment of the sacred in the profane. The authors start from the hypothesis of the revelation of an irreducible ontological structure of the human, which is the experience of the paradoxical limit between aformal and formal. Consequently, the endeavor of the authors is structured as a way of confirming their hypothesis through: 1) analysing the specificity of Eliade’s anti-reductionist approach as a historian and philosopher of religions and his openness towards anthropology; 2) the application of phenomenology and hermeneutics in order to show Eliade’s research scheme: facts - forms - symbolical structures - existential structures; 3) explaining the experience of the paradoxical limit between the sacred and the profane as a structure of the human; 4) inventorizing the human structures and the modes of subjectivation generated by the irreducibility of the experience of the paradoxical limit; 5) a brief account of the reinterpretation of the experience of the limit as a structure of the humanity in contemporary phenomenological writings. The authors consider that the experience of the paradoxical limit between the sacred and the profane, the miracle and the ordinary, is irreducible, and the proof is the interpretation of that limit in contemporary philosophical works in the form of: sovereignty (Georges Bataille), self-affection of Life (Michel Henry), or desire of the self through the other (Renaud Barbaras). The hypothesis of the experience of the limit or of the transformation of the man in the limit as a structure of the human is confirmed, as well as the interpretation of that limit by Eliade as coincidentia oppositorum.

Arheologii ale prezentului în Istoria literaturii române contemporane: 1990-2020

Authors: Adriana Stan

The paper disscusses the literary-historiographical enterprise of Mihai Iovănel’s History of the Contemporary Romanian Literature: 1990-2000 as a turning point in respect with the twofold aesthetic and ethnocentric dominant of this genre in Romania, whose tradition that is traceable back to the pos-Thaw period prolongs into postcommunism. Iovănel’s self-claimed ”postmarxist” approach puts under a deconstructive lens the ideological embededness of the Romanian literary space, whose postcommunist avatar still owes largely to its communist heritage by its indebtment to dominantly conservative and right ideologies. It also switches the analytical focus of the literary history towards the material (i.e. social, institutional, and so forth) conditions of the literary production per se. My essays develops two basic critical points derived from the above-mentioned premises. One concerns the topicality of the literary historian’s leftist vantage point in the view of deciphering long-engrained myths of the national literature and of its criticism, thus providing a clearer, more realist picture of its concrete phenomena. The other debates the utility and eventual drawbacks of the retrospective analysis the historian is forced to assume in order to achieve his deconstructive task, a retrospection that pushes his investigation way behind the confines of ”contemporaneity”.

Interleaving Automatic Segmentation and Expert Opinion for Retinal Conditions

Authors: Adrian Petru Groza

Optical coherence tomography (OCT) has become the leading diagnostic tool in modern ophthalmology. We are interested here in developing a support tool for the segmentation of retina layers. The proposed method relies on graph theory and geodesic distance. As each retina layer is characterised by different features, the proposed method interleaves various gradients during detection, such as horizontal and vertical gradients or open-closed gradients. The method was tested on a dataset of 750 OCT B-Scan Spectralis provided by the Ophthalmology Department of the County Emergency Hospital Cluj-Napoca. The method has smaller signed error on layers B1, B7 and B8, with the highest value of 0.43 pixels. The average value of signed error on all layers is -1.99 ± 1.14 px. The average value for mean absolute error is 2.60 ± 0.95 px. Since the target is a support tool for the human agent, the ophthalmologist can intervene after each automatic step. Human intervention includes validation or fine tuning of the automatic segmentation. In line with design criteria advocated by explainable artificial intelligence (XAI) and human-centered AI, this approach gives more control and transparency as well as more of a global perspective on the segmentation process.

Reliable Learning with PDE-Based CNNs and DenseNets for Detecting COVID-19, Pneumonia, and Tuberculosis from Chest X-Ray Images

Authors: Adrian Petru Groza, Anca Nicoleta Mărginean, Radu-Răzvan Slăvescu

It has recently been shown that the interpretation by partial differential equations (PDEs) of a class of convolutional neural networks (CNNs) supports definition of architectures such as parabolic and hyperbolic networks. These networks have provable properties regarding the stability against the perturbations of the input features. Aiming for robustness, we tackle the problem of detecting changes in chest X-ray images that may be suggestive of COVID-19 with parabolic and hyperbolic CNNs and with domain-specific transfer learning. To this end, we compile public data on patients diagnosed with COVID-19, pneumonia, and tuberculosis, along with normal chest X-ray images. The negative impact of the small number of COVID-19 images is reduced by applying transfer learning in several ways. For the parabolic and hyperbolic networks, we pretrain the networks on normal and pneumonia images and further use the obtained weights as the initializers for the networks to discriminate between COVID-19, pneumonia, tuberculosis, and normal aspects. For DenseNets, we apply transfer learning twice. First, the ImageNet pretrained weights are used to train on the CheXpert dataset, which includes 14 common radiological observations (e.g., lung opacity, cardiomegaly, fracture, support devices). Then, the weights are used to initialize the network which detects COVID-19 and the three other classes. The resulting networks are compared in terms of how well they adapt to the small number of COVID-19 images. According to our quantitative and qualitative analysis, the resulting networks are more reliable compared to those obtained by direct training on the targeted dataset.

Agents that argue and explain classifications of retinal conditions

Authors: Adrian Petru Groza

Abstract Purpose: Expertise for auditing AI systems in medical domain is only now being accumulated. Conformity assessment procedures will require AI systems: 1) to be transparent, 2) not to rely decisions solely on algorithms, or iii) to include safety assurance cases in the documentation to facilitate technical audit. We are interested here in obtaining transparency in the case of machine learning (ML) applied to classification of retina conditions. High performance metrics achieved using ML has become common practice. However, in the medical domain, algorithmic decisions need to be sustained by explanations. We aim at building a support tool for ophthalmologists able to: i) explain algorithmic decision to the human agent by automatically extracting rules from the ML learned models; ii) include the ophthalmologist in the loop by formalising expert rules and including the expert knowledge in the argumentation machinery; iii) build safety cases by creating assurance argument patterns for each diagnosis. Methods: For the learning task, we used a dataset consisting of 699 OCT images: 126 Normal class, 210 with Diabetic Retinopathy (DR) and 363 with Age Related Macular Degeneration (AMD). The dataset contains patients from the Ophthalmology Department of the County Emergency Hospital of Cluj-Napoca. All ethical norms and procedures, including anonymisation, have been performed. We applied three machine learning algorithms: decision tree (DT), support vector machine (SVM) and artificial neural network (ANN). For each algorithm we automatically extract diagnosis rules. For formalising expert knowledge, we relied on the normative dataset [13]. For arguing be- tween agents, we used the Jason multi-agent platform. We assume different knowledge base and reasoning capabilities for each agent. The agents have their own Optical Coherence Tomography (OCT) images on which they apply a distinct machine learning algorithm. The learned model is used to extract diagnosis rules. With distinct learned rules, the agents engage in an argumentative process. The resolution of the debate outputs a diagnosis that is then explained to the ophthalmologist, by means of assurance cases. Results: For diagnosing the retina condition, our AI solution deals with the following three issues: First, the learned models are automatically translated into rules. These rules are then used to build an explanation by tracing the reasoning chain supporting the diagnosis. Hence, the proposed AI solution complies with the requirement that ”algorithmic decision should be explained to the human agent”. Second, the decision is not solely based on ML-algorithms. The proposed architecture includes expert knowledge. The diagnosis is taken based on exchanging arguments between ML-based algorithms and expert knowledge. The conflict resolution among arguments is verbalised, so that the ophthalmologist can supervise the diagnosis. Third, the assurance cases are generated to facilitate technical audit. The assurance cases structure the evidence among various safety goals such as: machine learning methodology, transparency, or data quality. For each dimension, the auditor can check the provided evidence against the current best practices or safety standards. Conclusion: We developed a multi-agent system for retina conditions in which algorithmic decisions are sustained by explanations. The proposed tool goes behind most software in medical domain that focuses only on performance metrics. Our approach helps the technical auditor to approve software in the medical domain. Interleaving knowledge extracted from ML-models with ex- pert knowledge is a step towards balancing the benefits of ML with explainability, aiming at engineering reliable medical applications.

A Puzzle-Based Dataset for Natural Language Inference

Authors: Adrian Petru Groza

We provide here a dataset for tasks related to natural language understanding and natural language inference. The dataset contains logical puzzles in natural language from three domains: comparing puzzles, knighs and knaves, and zebra puzzles. Each puzzle is associated with the entire set of atomic questions that can be generated based on the relations and individuals occurring in the text. For each question we provide the correct answer: entailment, contradiction or ambiguity. The answer's correctness is verified against theorem provers. Good puzzles have two properties: (i) each piece of information is necessary and (ii) no unnecessary information is provided. These properties make puzzles interesting candidates for machine comprehension tasks.

Natural language understanding for logical games

Authors: Adrian Petru Groza

We developed a system able to automatically solve logical puzzles in natural language. Our solution is composed by a parser and an inference module. The parser translates the text into first order logic (FOL), while the MACE4 model finder is used to compute the models of the given FOL theory. We also empower our software agent with the capability to provide Yes/No answers to natural language questions related to each puzzle. Moreover, in line with Explainalbe Artificial Intelligence (XAI), the agent can back its answer, providing a graphical representation of the proof. The advantage of using reasoning for Natural Language Understanding (NLU) instead of Machine learning is that the user can obtain an explanation of the reasoning chain. We illustrate how the system performs on various types of natural language puzzles, including 382 knights and knaves puzzles. These features together with the overall performance rate of 80.89\% makes the proposed solution an improvement upon similar solvers for natural language understanding in the puzzles domain.

An Integrated Framework for Traceability and Impact Analysis in Requirements Verification of Cyber–Physical Systems

Authors: Adrian Pop

In the field of model-based design of Cyber–Physical Systems (CPS), seamless traceability of the process, from requirements to models to simulation results, is becoming increasingly important. It can be used to support several activities such as variant handling, impact analysis, component reuse, software maintenance and evolution, verification, and validation. Despite the fact that the relevance of traceability in the model-based design of CPSs is well known, current tools that support traceability management are inadequate in practice. The lack of comprehensive whole-lifecycle systems engineering support in a single tool is one of the main causes of such ineffective traceability management, where traceability relationships between artifacts are still manually generated and maintained. This paper aims at presenting an approach and a prototype for automatically generating and maintaining the appropriate traceability links between heterogeneous artifacts ranging from requirement models, through design models, down to simulation and verification results throughout the product life cycle in model-based design of CPSs. A use case study is presented to validate and illustrate the proposed method and prototype.

Open Source Languages and Methods for Cyber-Physical System Development: Overview and Case Studies

Authors: Adrian Pop

Industrial cyber-physical system products interleave hardware, software, and communication components. System complexity is increasing simultaneously with increased demands on quality and shortened time-to-market. To effectively support the development of such systems, we present languages and tools for comprehensive integrated model-based development that cover major phases such as requirement analysis, design, implementation, and maintenance. The model-based approach raises the level of abstraction and allows to perform virtual prototyping by simulating and optimizing system models before building physical products. Moreover, open standards and open source implementations enable model portability, tool reuse and a broader deployment. In this paper we present a general overview of the available solutions with focus on Modelica/OpenModelica, Bloqqi, and Acumen. The paper presents contributions to these languages and environments, including symbolic-numeric modeling, requirement verification, code generation, model debugging, design optimization, graphical modeling, and variant handling with traceability, as well a general discussion and conclusions.

OpenModelica.jl: A modular and extensible Modelica compiler framework in Julia targeting ModelingToolkit.jl

Authors: Adrian Pop

This paper presents current work on our Modelica Compiler framework in Julia: OpenModelica.jl. We provide a brief overview of this novel framework and its features, and we also present the latest addition to the possible backend options. We target ModelingToolkit.jl (MTK), a framework for symbolic-numerical computation and scientific machine learning. We evaluated the performance of our new backend using the ScalableTestsuite, a benchmark suite for Modelica Compilers. In our experiment, we demonstrate that MTK can be used as a backend with competitive simulation performance. In addition, using the scientific machine learning features of the Modeling toolkit, we were able to approximate models in the ScalableTestsuite using surrogate techniques and how such techniques can be used to accelerate the solving of nonlinear algebraic loops during tearing. Based on our experiments, we propose using this new framework to automatically generate surrogate components of a Modelica model during the simulation to increase performance. The experimental work presented here provides one of the first investigations concerning the integration of the symbolic-numerical abilities of Julia within a Modelica tool.

New Method to Perform Data Reconciliation with OpenModelica and ThermoSysPro

Authors: Adrian Pop

Data reconciliation aims at improving the accuracy of measurements by reducing the effect of random errors in the data. This is achieved by introducing redundancies on the measured quantities in the form of constraints based on fundamental physical laws such as mass, momentum and energy balance equations. These constraints are called the auxiliary conditions. Modelica is an equational language that was conceived to express models based on first principle physics for the purpose of behavioral simulation. This paper shows how to reuse such models for the purpose of data reconciliation. The novelty is to automatically extract the auxiliary conditions from the Modelica model. Then the reconciled values are computed using a least square method constrained by the auxiliary conditions, as specified by the VDI 2048 standard. The new method has been implemented in OpenModelica. A simple example built with ThermoSysPro illustrates the method in detail.

Generating the Voice of the Interactive Virtual Assistant

Authors: Adriana Stan

This chapter introduces an overview of the current approaches for generating spoken content using text-to-speech synthesis (TTS) systems, and thus the voice of an Interactive Virtual Assistant (IVA). The overview builds upon the issues which make spoken content generation a non-trivial task, and introduces the two main components of a TTS system: text processing and acoustic modelling. It then focuses on providing the reader with the minimally required scientific details of the terminology and methods involved in speech synthesis, yet with sufficient knowledge so as to be able to make the initial decisions regarding the choice of technology for the vocal identity of the IVA. The speech synthesis methodologies’ description begins with the basic, easy to run, low-requirement rule-based synthesis, and ends up within the state-of-the-art deep learning landscape. To bring this extremely complex and extensive research field closer to commercial deployment, an extensive indexing of the readily and freely available resources and tools required to build a TTS system is provided. Quality evaluation methods and open research problems are, as well, highlighted at end of the chapter.

An objective evaluation of the effects of recording conditions and speaker characteristics in multi-speaker deep neural speech synthesis

Authors: Adriana Stan, Mircea Giurgiu

Multi-speaker spoken datasets enable the creation of text-to-speech synthesis (TTS) systems which can output several voice identities. The multi-speaker (MSPK) scenario also enables the use of fewer training samples per speaker. However, in the resulting acoustic model, not all speakers exhibit the same synthetic quality, and some of the voice identities cannot be used at all. In this paper we evaluate the influence of the recording conditions, speaker gender, and speaker particularities over the quality of the synthesised output of a deep neural TTS architecture, namely Tacotron2. The evaluation is possible due to the use of a large Romanian parallel spoken corpus containing over 81 hours of data. Within this setup, we also evaluate the influence of different types of text representations: orthographic, phonetic, and phonetic extended with syllable boundaries and lexical stress markings. We evaluate the results of the MSPK system using the objective measures of equal error rate (EER) and word error rate (WER), and also look into the distances between natural and synthesised t-SNE projections of the embeddings computed by an accurate speaker verification network. The results show that there is indeed a large correlation between the recording conditions and the speaker’s synthetic voice quality. The speaker gender does not influence the output, and that extending the input text representation with syllable boundaries and lexical stress information does not equally enhance the generated audio across all speaker identities. The visualisation of the t-SNE projections of the natural and synthesised speaker embeddings show that the acoustic model shifts some of the speakers’ neural representation, but not all of them. As a result, these speakers have lower performances of the output speech.

An evaluation of word-level confidence estimation for end-to-end automatic speech recognition

Authors: Adriana Stan

Quantifying the confidence (or conversely the uncertainty) of a prediction is a highly desirable trait of an automatic system, as it improves the robustness and usefulness in downstream tasks. In this paper we investigate confidence estimation for end-to-end automatic speech recognition (ASR). Previous work has addressed confidence measures for lattice-based ASR, while current machine learning research mostly focuses on confidence measures for unstructured deep learning. However, as the ASR systems are increasingly being built upon deep end-to-end methods, there is little work that tries to develop confidence measures in this context. We fill this gap by providing an extensive benchmark of popular confidence methods on four well-known speech datasets. There are two challenges we overcome in adapting existing methods: working on structured data (sequences) and obtaining confidences at a coarser level than the predictions (words instead of tokens). Our results suggest that a strong baseline can be obtained by scaling the logits by a learnt temperature, followed by estimating the confidence as the negative entropy of the predictive distribution and, finally, sum pooling to aggregate at word level.

Speaker disentanglement in video-to-speech conversion

Authors: Adriana Stan

The task of video-to-speech aims to translate silent video of lip movement to its corresponding audio signal. Previous approaches to this task are generally limited to the case of a single speaker, but a method that accounts for multiple speakers is desirable as it allows to i) leverage datasets with multiple speakers or few samples per speaker; and ii) control speaker identity at inference time. In this paper, we introduce a new video-to-speech architecture and explore ways of extending it to the multi-speaker scenario: we augment the network with an additional speaker-related input, through which we feed either a discrete identity or a speaker embedding. Interestingly, we observe that the visual encoder of the network is capable of learning the speaker identity from the lip region of the face alone. To better disentangle the two inputs -- linguistic content and speaker identity -- we add adversarial losses that dispel the identity from the video embeddings. To the best of our knowledge, the proposed method is the first to provide important functionalities such as i) control of the target voice and ii) speech synthesis for unseen identities over the state-of-the-art, while still maintaining the intelligibility of the spoken output.

Speaker verification-derived loss and data augmentation for DNN-based multispeaker speech synthesis

Authors: Adriana Stan, Mircea Giurgiu

Building multispeaker neural network-based text-to-speech synthesis systems commonly relies on the availability of large amounts of high quality recordings from each speaker and conditioning the training process on the speaker's identity or on a learned representation of it. However, when little data is available from each speaker, or the number of speakers is limited, the multispeaker TTS can be hard to train and will result in poor speaker similarity and naturalness. In order to address this issue, we explore two directions: forcing the network to learn a better speaker identity representation by appending an additional loss term; and augmenting the input data pertaining to each speaker using waveform manipulation methods. We show that both methods are efficient when evaluated with both objective and subjective measures. The additional loss term aids the speaker similarity, while the data augmentation improves the intelligibility of the multispeaker TTS system.

Distributed Ledger Technology Review and Decentralized Applications Development Guidelines

Authors: Tudor Cioara, Anghel Ionuț

The Distributed Ledger Technology (DLT) provides an infrastructure for developing decentralized applications with no central authority for registering, sharing, and synchronizing transactions on digital assets. In the last years, it has drawn high interest from the academic community, technology developers, and startups mostly by the advent of its most popular type, blockchain technology. In this paper, we provide a comprehensive overview of DLT analyzing the challenges, provided solutions or alternatives, and their usage for developing decentralized applications. We define a three-tier based architecture for DLT applications to systematically classify the technology solutions described in over 100 papers and startup initiatives. Protocol and Network Tier contains solutions for digital assets registration, transactions, data structure, and privacy and business rules implementation and the creation of peer-to-peer networks, ledger replication, and consensus-based state validation. Scalability and Interoperability Tier solutions address the scalability and interoperability issues with a focus on blockchain technology, where they manifest most often, slowing down its large-scale adoption. The paper closes with a discussion on challenges and opportunities for developing decentralized applications by providing a multi-step guideline for decentralizing the design and implementation of traditional systems.

Blockchain based decentralized local energy flexibility market

Authors: Tudor Cioara, Anghel Ionuț, Dan Mitrea

Large-scale deployment of renewable energy sources brings new challenges for smart grid management requiring the development of decentralized solutions and active participation of prosumer and non-grid-owned assets. Local energy flexibility markets can help in monitoring energy flows, motivate changes in prosumers’ energy supply and demand, achieving local energy balance, and optimization of electricity flows. In this paper, we propose a blockchain-based decentralized energy flexibility market enabling small-scale prosumers to trade in a peer-to-peer fashion their flexibility in terms of load modulation concerning the baseline energy profiles. We have defined an energy flexibility token for digitizing the flexibility of prosumers allowing to be traded on the market as an asset and self-enforcing smart contracts for decentralized market operation including functions such as the placement of flexibility bids/offers, trading session management, or energy and financial settlement of energy flexibility transactions. For matching the flexibility bids and offers, a solution based on a greedy heuristic and a bipartite graph is proposed for minimizing the number of flexibility transactions and reducing the blockchain-associated costs, while Oracles are used to assure its secure integration with the blockchain. The blockchain-based flexibility market was validated with the help of the Terni city Distribution System Operator, showing promising results in enabling the self-consumption of renewable energy generated in a small scale urban micro-grid considering live energy monitoring data, and in assuring the local balancing of the demand side in a simulated environment considering many market participants and historical energy data.

Blockchain-Based Decentralized Virtual Power Plants of Small Prosumers

Authors: Tudor Cioara, Anghel Ionuț, Dan Mitrea

The deployment of small-scale renewable energy sources will transform the management of energy grids towards more decentralized solutions in which the prosumers will have a more active role. Regulatory and market barriers are driving the implementation of virtual aggregation models in which the small-scale prosumers work together on a larger scale to gain benefits that could not be obtained on an individual basis. In this paper, we propose to use public blockchain and self-enforcing smart contracts to construct Virtual Power Plants (VPPs) of prosumers to provide energy services. A model has been defined for capturing the prosumer level constraints in terms of available energy profiles and energy service requirements enabling their optimal aggregation in hierarchical structures. A lightweight decentralized solution for VPPs construction is implemented using smart contracts enabling its efficient running on the public blockchain. Smart contracts are encoding the model constraints and are defining functionalities for prosumers to initiate or join a VPP implementing the complete chain of Offer-Operate-Measure-Remunerate actions. The VPP will be managed on top of a distributed ledger technology offering decentralized functionality for tracking and validating the delivery of energy based on the blockchain transactions and for energy and financial settlement, the remuneration being done according to the amount of energy provided by individual prosumers. Experimental results show that the proposed solution runs successfully on the public blockchain with good execution time and can address Balancing Responsible Party requests for additional generation. The overhead in terms of gas consumption and transactional throughput stays within reasonable boundaries.

Heating Homes with Servers: Workload Scheduling for Heat Reuse in Distributed Data Centers

Authors: Tudor Cioara, Anghel Ionuț

Data centers consume lots of energy to execute their computational workload and generate heat that is mostly wasted. In this paper, we address this problem by considering heat reuse in the case of a distributed data center that features IT equipment (i.e., servers) installed in residential homes to be used as a primary source of heat. We propose a workload scheduling solution for distributed data centers based on a constraint satisfaction model to optimally allocate workload on servers to reach and maintain the desired home temperature setpoint by reusing residual heat. We have defined two models to correlate the heat demand with the amount of workload to be executed by the servers: a mathematical model derived from thermodynamic laws calibrated with monitored data and a machine learning model able to predict the amount of workload to be executed by a server to reach a desired ambient temperature setpoint. The proposed solution was validated using the monitored data of an operational distributed data center. The server heat and power demand mathematical model achieve a correlation accuracy of 11.98% while in the case of machine learning models, the best correlation accuracy of 4.74% is obtained for a Gradient Boosting Regressor algorithm. Also, our solution manages to distribute the workload so that the temperature setpoint is met in a reasonable time, while the server power demand is accurately following the heat demand.

Designing an eHealth Coaching Solution to Improve Transitional Care of Seniors’ with Heart Failure: End-User Needs

Authors: Tudor Cioara, Anghel Ionuț

Based on scientific studies, heart failure is the principal cause of hospitalization among seniors. More than 50% of elderly with heart failure are readmitted to hospital within six months. Readmission is linked with poor compliance with medical treatment and recommendations, emphasizing the need for a tool to help seniors better comply with post-discharge measures. The goal of this study was to identify end-user needs for the development of a coaching solution aiming to support elderly patients but also formal and informal caregivers. End-user needs were identified through interviews with the three end-user profiles: seniors with heart failure and formal and informal caregivers. The results present six categories of needs: daily treatment follow-up; healthcare network communication; transfer of information; synchronization with current digital tools; information access; and psychosocial support. The identified needs will help to develop an eHealth solution to improve care management and coaching after discharge.

An Overview of Digital Twins Application Domains in Smart Energy Grid

Authors: Tudor Cioara, Anghel Ionuț

The Digital Twins offer promising solutions for smart grid challenges related to the optimal operation, management, and control of energy assets, for safe and reliable distribution of energy. These challenges are more pressing nowadays than ever due to the large-scale adoption of distributed renewable resources at the edge of the grid. The digital twins are leveraging technologies such as the Internet of Things, big data analytics, machine learning, and cloud computing, to analyze data from different energy sensors, view and verify the status of physical energy assets and extract useful information to predict and optimize the assets performance. In this paper, we will provide an overview of the Digital Twins application domains in the smart grid while analyzing existing the state of the art literature. We have focused on the following application domains: energy asset modeling, fault and security diagnosis, operational optimization, and business models. Most of the relevant literature approaches found are published in the last three years showing that the domain of Digital Twins application in smart grid is hot and gradually developing. Anyway, there is no unified view on the Digital Twins implementation and integration with energy management processes, thus, much work still needs to be done to understand and automatize the smart grid management.

A Lockable ERC20 Token for Peer to Peer Energy Trading

Authors: Tudor Cioara, Anghel Ionuț, Dan Mitrea

In this paper, we address the digitization of physical assets using blockchain technology focusing on energy and peer-to-peer trading on decentralized energy markets. Because they are forward markets and operate on a day-ahead timeline, the energy transactions are settled only at the movement of energy delivery. Having the option of locking the energy tokens by a third-party escrow becomes a highly desirable feature. Thus, we define a Lockable ERC20 token that provides the option for an owner to lock some of its tokens using smart contracts. A time-lock and an escrow party account or smart contract can be specified allowing the tokens to be unlocked when certain business conditions are met. For validation purposes, we have considered a peer-to-peer energy trading scenario in which the Lockable ERC20 token was used to digitize the surplus of energy of prosumers. In this case, the energy tokens committed in blockchain transactions are successfully locked up until the actual delivery of energy, the settlement considering the monitored data of energy meters.

Answer Counting Under Guarded TGDs.

Authors: Cristina Feier

We study the complexity of answer counting for ontology-mediated queries and for querying under constraints, considering conjunctive queries and unions thereof (UCQs) as the query language and guarded TGDs as the ontology and constraint language, respectively. Our main result is a classification according to whether answer counting is fixed-parameter tractable (FPT), W[1]-equivalent, #W[1]-equivalent, #W[2]-hard, or #A[2]-equivalent, lifting a recent classification for UCQs without ontologies and constraints due to Dell et al. [Holger Dell et al., 2019]. The classification pertains to various structural measures, namely treewidth, contract treewidth, starsize, and linked matching number. Our results rest on the assumption that the arity of relation symbols is bounded by a constant and, in the case of ontology-mediated querying, that all symbols from the ontology and query can occur in the data (so-called full data schema). We also study the meta-problems for the mentioned structural measures, that is, to decide whether a given ontology-mediated query or constraint-query specification is equivalent to one for which the structural measure is bounded.

Characterising Fixed Parameter Tractability for Query Evaluation Over Guarded TGDs

Authors: Cristina Feier

We consider the parameterized complexity of evaluating Ontology Mediated Queries (OMQ) based on Guarded TGDs (GTGD) and Unions of Conjunctive Queries, in the case where relational symbols have unrestricted arity and where the parameter is the size of the OMQ. We establish exact criteria for fixed-parameter tractable (fpt) evaluation of recursively enumerable (r.e.) classes of such OMQs (under the widely held Exponential Time Hypothesis). One of the main technical tools introduced in the paper is an fpt-reduction from deciding parameterized uniform CSPs to parameterized OMQ evaluation. The reduction preserves measures known to be essential for classifying r.e. classes of parameterized uniform CSPs: submodular width (according to the well known result of Marx for unrestricted-arity schemas) and treewidth (according to the well known result of Grohe for bounded-arity schemas). As such, it can be employed to obtain hardness results for evaluation of r.e. classes of parameterized OMQs based on GTGD both in the unrestricted and in the bounded arity case. Previously, for bounded arity schemas, this has been tackled using a technique requiring full introspection into the construction employed by Grohe.

Efficiency of Query Evaluation Under Guarded TGDs: The Unbounded Arity Case.

Authors: Cristina Feier

The paper analyzes the parameterized complexity of evaluating Ontology Mediated Queries (OMQs) based on Guarded TGDs (GTGDs) and Unions of Conjunctive Queries (UCQs), in the setting where relational symbols might have unbounded arity and where the parameter is the size of the OMQ. It establishes exact criteria for fixed-parameter tractability (fpt) evaluation of recursively enumerable classes of such OMQs (under the widely held Exponential Time Hypothesis). One of the main technical tools introduced in the paper is an fpt-reduction from deciding parameterized uniform CSPs to parameterized OMQ evaluation. A fundamental feature of the reduction is preservation of measures which are known to be essential for classifying classes of parameterized uniform CSPs: submodular width (according to the well known result of Marx for unbounded-arity schemas) and treewidth (according to the well known result of Grohe for bounded-arity schemas). As such, the reduction can be employed to obtain hardness results for evaluation of classes of parameterized OMQs both in the unbounded and in the bounded arity case. Previously, in the case of bounded arity schemas, this has been tackled using a technique requiring full introspection into the construction employed by Grohe.

ObserveNet Control: A Vision-Dynamics Learning Approach to Predictive Control in Autonomous Vehicles

Authors: Lucian Bușoniu

A key component in autonomous driving is the ability of the self-driving car to understand, track and predict the dynamics of the surrounding environment. Although there is significant work in the area of object detection, tracking and observations prediction, there is no prior work demonstrating that raw observations prediction can be used for motion planning and control. In this paper, we propose ObserveNet Control, which is a vision-dynamics approach to the predictive control problem of autonomous vehicles. Our method is composed of a i) deep neural network able to confidently predict future sensory data on a time horizon of up to 10 s and ii) a temporal planner designed to compute a safe vehicle state trajectory based on the predicted sensory data. Given the vehicle's historical state and sensing data in the form of Lidar point clouds, the method aims to learn the dynamics of the observed driving environment in a self-supervised manner, without the need to manually specify training labels. The experiments are performed both in simulation and real-life, using CARLA and RovisLab's AMTU mobile platform as a 1:4 scaled model of a car. We evaluate the capabilities of ObserveNet Control in aggressive driving contexts, such as overtaking maneuvers or side cut-off situations, while comparing the results with a baseline Dynamic Window Approach (DWA) and two state-of-the-art imitation learning systems, that is, Learning by Cheating (LBC) and World on Rails (WOR).

Fast constrained nonlinear model predictive control for implementation on microcontrollers

Authors: Lucian Bușoniu

Model predictive control (MPC) is based on the systematic resolution of an online optimization problem at each time step. In practice, the computation cost is often very high, especially for the non-linear case under constraints, thus complicating the application of MPC to real-time systems. This paper proposes to improve the non-linear quadratic dynamic matrix control (NLQDMC) algorithm for MPC by solving constrained optimization problems only when necessary, and defaulting to the unconstrained solution whenever possible. The new algorithm is called fast NLQDMC (FNLQDMC) and is applied to the control of a nonlinear system comprised of converter and a DC machine, and implemented in a microcontroller board. The results obtained show that, depending to the setpoint profiles, this algorithm saves more than 64% computation of the constrained problem compared to the conventional NLQDMC, while keeping identical performance in terms of setpoint tracking and constraint satisfactions.

Stable Near-Optimal Control of Nonlinear Switched Discrete-Time Systems: An Optimistic Planning-Based Approach

Authors: Lucian Bușoniu

Originating in the artificial intelligence literature, optimistic planning (OP) is an algorithm that generates near-optimal control inputs for generic nonlinear discrete-time systems whose input set is finite. This technique is, therefore, relevant for the near-optimal control of nonlinear switched systems for which the switching signal is the control, and no continuous input is present. However, OP exhibits several limitations, which prevent its desired application in a standard control engineering context, as it requires, for instance, that the stage cost takes values in$[0, 1]$, an unnatural prerequisite, and that the cost function is discounted. In this article, we modify OP to overcome these limitations, and we call the new algorithm${\rm OP}_{\text{min}}$. We then analyze${\rm OP}_{\text{min}}$under general stabilizability and detectability assumptions on the system and the stage cost. New near-optimality and performance guarantees for${\rm OP}_{\text{min}}$are derived, which have major advantages compared to those originally given for OP. We also prove that a system whose inputs are generated by${\rm OP}_{\text{min}}$in a receding-horizon fashion exhibits stability properties. As a result,${\rm OP}_{\text{min}}$provides a new tool for the near-optimal, stable control of nonlinear switched discrete-time systems for generic cost functions.

Exploiting homogeneity for the optimal control of discrete-time systems: application to value iteration

Authors: Lucian Bușoniu

To investigate solutions of (near-)optimal control problems, we extend and exploit a notion of homogeneity recently proposed in the literature for discrete-time systems. Assuming the plant dynamics is homogeneous, we first derive a scaling property of its solutions along rays provided the sequence of inputs is suitably modified. We then consider homogeneous cost functions and reveal how the optimal value function scales along rays. This result can be used to construct (near-)optimal inputs on the whole state space by only solving the original problem on a given compact manifold of a smaller dimension. Compared to the related works of the literature, we impose no conditions on the homogeneity degrees. We demonstrate the strength of this new result by presenting a new approximate scheme for value iteration, which is one of the pillars of dynamic programming. The new algorithm provides guaranteed lower and upper estimates of the true value function at any iteration and has several appealing features in terms of reduced computation. A numerical case study is provided to illustrate the proposed algorithm.

Path-aware optimistic optimization for a mobile robot

Authors: Lucian Bușoniu

We consider problems in which a mobile robot samples an unknown function defined over its operating space, so as to find a global optimum of this function. The path travelled by the robot matters, since it influences energy and time requirements. We consider a branch-and-bound algorithm called deterministic optimistic optimization, and extend it to the path-aware setting, obtaining path-aware optimistic optimization (OOPA). In this new algorithm, the robot decides how to move next via an optimal control problem that maximizes the long-term impact of the robot trajectory on lowering the upper bound, weighted by bound and function values to focus the search on the optima. An online version of value iteration is used to solve an approximate version of this optimal control problem. OOPA is evaluated in extensive experiments in two dimensions, where it does better than path-unaware and local-optimization baselines.

Spectral Normalisation for Deep Reinforcement Learning: an Optimisation Perspective

Authors: Lucian Bușoniu, Razvan Pascanu

Most of the recent deep reinforcement learning advances take an RL-centric perspective and focus on refinements of the training objective. We diverge from this view and show we can recover the performance of these developments not by changing the objective, but by regularising the value-function estimator. Constraining the Lipschitz constant of a single layer using spectral normalisation is sufficient to elevate the performance of a Categorical-DQN agent to that of a more elaborated \rainbow{} agent on the challenging Atari domain. We conduct ablation studies to disentangle the various effects normalisation has on the learning dynamics and show that is sufficient to modulate the parameter updates to recover most of the performance of spectral normalisation. These findings hint towards the need to also focus on the neural component and its learning dynamics to tackle the peculiarities of Deep Reinforcement Learning.

Pedestrian Street-Cross Action Recognition in Monocular Far Infrared Sequences

Authors: Sergiu Nedevschi, Raluca Didona Brehar, Mircea Paul Mureșan, Tiberiu Marița, Cristian Vancea

The early recognition and understanding of the actions performed by pedestrians in traffic scenes leads to an anticipation of pedestrian intentions in advance and helps in the process of collision warning and avoidance in the context of autonomous vehicles. An environment with low visibility conditions such as night-time, fog, heavy rain or smoke increases the number of difficult situations in traffic. A complete and original model for assessing if a pedestrian is engaged in a street cross action using only infrared monocular scene perception is proposed in this paper. The assessment of a street cross action is done by the time series analysis of features like: pedestrian motion, position of pedestrians with respect to the drivable area and their distance with respect to the ego-vehicle. The extraction of these features emerges from the combination of a deep learning based pedestrian detector with an original tracking algorithm, a semantic segmentation of the road surface and a time series long-short term memory network based action recognition. In order to validate the proposed method we introduce a new dataset named CROSSIR. It is formed of pedestrian annotations, action annotations and semantic labels for the road. The CROSSIR dataset is suitable for several common computer vision algorithms: (1) pedestrian detection and tracking algorithms because each pedestrian has a unique identifier over the frames in which it appears; (2) pedestrian action recognition; (3) semantic segmentation of the road pixels in the infrared image.

Hepatocellular Carcinoma Automatic Diagnosis within CEUS and B-Mode Ultrasound Images Using Advanced Machine Learning Methods

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea, Stelian Brad

Hepatocellular Carcinoma (HCC) is the most common malignant liver tumor, being present in 70% of liver cancer cases. It usually evolves on the top of the cirrhotic parenchyma. The most reliable method for HCC diagnosis is the needle biopsy, which is an invasive, dangerous method. In our research, specific techniques for non-invasive, computerized HCC diagnosis are developed, by exploiting the information from ultrasound images. In this work, the possibility of performing the automatic diagnosis of HCC within B-mode ultrasound and Contrast-Enhanced Ultrasound (CEUS) images, using advanced machine learning methods based on Convolutional Neural Networks (CNN), was assessed. The recognition performance was evaluated separately on B-mode ultrasound images and on CEUS images, respectively, as well as on combined B-mode ultrasound and CEUS images. For this purpose, we considered the possibility of combining the input images directly, performing feature level fusion, then providing the resulted data at the entrances of representative CNN classifiers. In addition, several multimodal combined classifiers were experimented, resulted by the fusion, at classifier, respectively, at the decision levels of two different branches based on the same CNN architecture, as well as on different CNN architectures. Various combination methods, and also the dimensionality reduction method of Kernel Principal Component Analysis (KPCA), were involved in this process. These results were compared with those obtained on the same dataset, when employing advanced texture analysis techniques in conjunction with conventional classification methods and also with equivalent state-of-the-art approaches. An accuracy above 97% was achieved when our new methodology was applied.

Robust Data Association Using Fusion of Data-Driven and Engineered Features for Real-Time Pedestrian Tracking in Thermal Images

Authors: Sergiu Nedevschi, Mircea Paul Mureșan, Radu Gabriel Dănescu

Object tracking is an essential problem in computer vision that has been extensively researched for decades. Tracking objects in thermal images is particularly difficult because of the lack of color information, low image resolution, or high similarity between objects of the same class. One of the main challenges in multi-object tracking, also referred to as the data association problem, is finding the correct correspondences between measurements and tracks and adapting the object appearance changes over time. We addressed this challenge of data association for thermal images by proposing three contributions. The first contribution consisted of the creation of a data-driven appearance score using five Siamese Networks, which operate on the image detection and on parts of it. Secondly, we engineered an original edge-based descriptor that improves the data association process. Lastly, we proposed a dataset consisting of pedestrian instances that were recorded in different scenarios and are used for training the Siamese Networks. The data-driven part of the data association score offers robustness, while feature engineering offers adaptability to unknown scenarios and their combination leads to a more powerful tracking solution. Our approach had a running time of 25 ms and achieved an average precision of 86.2% on publicly available benchmarks, containing real-world scenarios, as shown in the evaluation section.

Pedestrian identification in infrared and visible images based on pose keypoints matching

Authors: Sergiu Nedevschi, Raluca Didona Brehar, Tiberiu Marița

Abstract The identification of persons in multi-spectral images that are not spatially aligned is a challenging process. A correct identification can improve the pedestrian detection task for machine vision applications. In this context we propose a person identification mechanism able to correctly find the same person in infrared and visible images. The main contribution of the paper is the keypoint based matcher that uses a deep learning based solution for finding relevant human pose keypoints and a local neighbour search for extracting the best candidates for person identification. For each of the two images, infrared and color we first perform pedestrian detection. Next, on each detected instance we extract the relevant keypoints for shoulders, hands and legs. A matching algorithm between the extracted keypoints is proposed in order to perform the identification of persons in the two images. We obtain an identification accuracy of 76% for pedestrians that have a medium height with respect to the image dimensions, and have a small occlusion degree.

Deep Learning for Short-Term Load Forecasting—Industrial Consumer Case Study

Authors: Vasile Topa

In the current trend of consumption, electricity consumption will become a very high cost for the end-users. Consumers acquire energy from suppliers who use short, medium, and long-term forecasts to place bids in the power market. This study offers a detailed analysis of relevant literature and proposes a deep learning methodology for forecasting industrial electric usage for the next 24 h. The hourly load curves forecasted are from a large furniture factory. The hourly data for one year is split into training (80%) and testing (20%). The algorithms use the previous two weeks of hourly consumption and exogenous variables as input in the deep neural networks. The best results prove that deep recurrent neural networks can retain long-term dependencies in high volatility time series. Gated recurrent units (GRU) obtained the lowest mean absolute percentage error of 4.82% for the testing period. The GRU improves the forecast by 6.23% compared to the second-best algorithm implemented, a combination of GRU and Long short-term memory (LSTM). From a practical perspective, deep learning methods can automate the forecasting processes and optimize the operation of power systems.

High Frequency Analysis and Optimization of Planar Spiral Inductors Used in Microelectronic Circuits

Authors: Vasile Topa, Călin Munteanu

This paper deals with high frequency analysis of spiral inductors, used in microelectronics circuits, to optimize their configuration. Software developed, designed, and implemented by the authors for nano and micrometre spiral inductor high frequency analysis, named ABSIF, is presented in this paper. ABSIF determines the inductance, quality factor, and electrical parameters for square, hexagonal, octagonal, and circular spiral inductors and their configuration optimization for energy efficiency. ABSIF is a good tool for spiral inductor design optimization in high frequency applications and takes into account the imposed technological limits and/or the designers’ constraints. A set of spiral inductors are considered and analysed for high frequency values using ABSIF, and the results are presented in the paper. The validation of ABSIF was completed by comparing the results with those obtained using a similar commercial software, Sonnet LiteTM, which is dedicated to high frequency electromagnetic analysis.

Analysis for Non-Residential Short-Term Load Forecasting Using Machine Learning and Statistical Methods with Financial Impact on the Power Market

Authors: Vasile Topa

Short-term load forecasting predetermines how power systems operate because electricity production needs to sustain demand at all times and costs. Most load forecasts for the non-residential consumers are empirically done either by a customer’s employee or supplier personnel based on experience and historical data, which is frequently not consistent. Our objective is to develop viable and market-oriented machine learning models for short-term forecasting for non-residential consumers. Multiple algorithms were implemented and compared to identify the best model for a cluster of industrial and commercial consumers. The article concludes that the sliding window approach for supervised learning with recurrent neural networks can learn short and long-term dependencies in time series. The best method implemented for the 24 h forecast is a Gated Recurrent Unit (GRU) applied for aggregated loads over three months of testing data resulted in 5.28% MAPE and minimized the cost with 5326.17 € compared with the second-best method LSTM. We propose a new model to evaluate the gap between evaluation metrics and the financial impact of forecast errors in the power market environment. The model simulates bidding on the power market based on the 24 h forecast and using the Romanian day-ahead market and balancing prices through the testing dataset.

IEEE Transactions on Intelligent Vehicles

Authors: Sergiu Nedevschi

The IEEE TRANSACTIONS ON INTELLIGENT VEHICLES (T-IV) publishes peer-reviewed articles that provide innovative research concepts and application results, report signifi cant theoretical fi ndings and application case studies, and raise awareness of pressing research and application challenges in areas of intelligent vehicles in a roadway environment, and in particular in automated and vehicles. The T-IV focuses on providing critical information to the intelligent vehicle community, serving as a dissemination vehicle for IEEE ITS Society members and the others to learn the state of the art development and progress on research and applications in the fi eld of intelligent vehicles. Member copies are for personal use only.

Detecting fake news for the new coronavirus by reasoning on the Covid-19 ontology

Authors: Adrian Petru Groza

In the context of the Covid-19 pandemic, many were quick to spread deceptive information. I investigate here how reasoning in Description Logics (DLs) can detect inconsistencies between trusted medical sources and not trusted ones. The not-trusted information comes in natural language (e.g. "Covid-19 affects only the elderly"). To automatically convert into DLs, I used the FRED converter. Reasoning in Description Logics is then performed with the Racer tool.

The OpenModelica Integrated Environment for Modeling, Simulation, and Model-Based Development

Authors: Adrian Pop

OpenModelica is a unique large-scale integrated open-source Modelica-and FMI-based modeling, simulation, optimization, model-based analysis and development environment.Moreover, the OpenModelica environment provides a number of facilities such as debugging; optimization; visualization and 3D animation; web-based model editing and simulation; scripting from Modelica, Python, Julia, and Matlab; efficient simulation and co-simulation of FMI-based models; compilation for embedded systems; Modelica-UML integration; requirement verification; and generation of parallel code for multi-core architectures.The environment is based on the equation-based object-oriented Modelica language and currently uses the MetaModelica extended version of Modelica for its model compiler implementation.This overview paper gives an up-to-date description of the capabilities of the system, short overviews of used open source symbolic and numeric algorithms with pointers to published literature, tool integration aspects, some lessons learned, and the main vision behind its development.

Traceability in the Model-based Design of Cyber-Physical Systems

Authors: Adrian Pop

Design, development, and analysis of complex Cyber Physical Systems (CPSs) using models involves a collaboration of expertise from different engineering domains.Heterogeneous artefacts are generated, often using different lifecycle modeling languages and simulation tools.Capturing the traceability information among these artefacts can be used to support several activities such as requirements tracing, impact analysis of change requests, verification, validation, and documentation.However, creating trace links among these heterogeneous artefacts is challenging as different tools in the development lifecycle are usually disparate and there is no precise semantic in the terminology used between requirement engineers, verification engineers, and system modelers.In this paper, we present a linked data-based approach to capture traceability information and create trace links that relate heterogeneous artefacts in the model-based design process of CPSs through a standardized interface and format using OSLC.This enables artefacts from different tools to be connected and queried through a standardized interface and format.A practical prototype system for supporting traceability is designed through integration with the INTO-CPS toolchain of CPS design.The traceability data is stored in Neo4j graph database which can be queried for generating various reports such as impact analysis, variant handling, etc.

Towards an Open-Source Modelica Compiler in Julia

Authors: Adrian Pop

Recently the Julia language has become an option for scientific computing.As of 2020, efforts exist to provide libraries that emulate the equation-based modeling features provided by Modelica or otherwise provide such functionality in Julia.The issue with these approaches is that investment in standardization and libraries would be lost unless standard-compliance is guaranteed.We believe that it is possible to combine features from both by implementing such a compiler in Julia.We argue that this approach would open additional opportunities.One such being the handling of variable structure systems (VSS) within the framework of a Modelica standard-compliant compiler.The other being a proposed compiler architecture reminiscent of LLVM for equation-based objectoriented languages.Using the OpenModelica Compiler as a baseline, we verified the fidelity of our implementation by simulating a selected set of models.While there are performance penalties, we argue that improvements to the frontend would mitigate these issues.

RECOApy: Data Recording, Pre-Processing and Phonetic Transcription for End-to-End Speech-Based Applications

Authors: Adriana Stan

Deep learning enables the development of efficient end-to-end speech processing applications while bypassing the need for expert linguistic and signal processing features. Yet, recent studies show that good quality speech resources and phonetic transcription of the training data can enhance the results of these applications. In this paper, the RECOApy tool is introduced. RECOApy streamlines the steps of data recording and pre-processing required in end-to-end speech-based applications. The tool implements an easy-to-use interface for prompted speech recording, spectrogram and waveform analysis, utterance-level normalisation and silence trimming, as well grapheme-to-phoneme conversion of the prompts in eight languages: Czech, English, French, German, Italian, Polish, Romanian and Spanish. The grapheme-to-phoneme (G2P) converters are deep neural network (DNN) based architectures trained on lexicons extracted from the Wiktionary online collaborative resource. With the different degree of orthographic transparency, as well as the varying amount of phonetic entries across the languages, the DNN's hyperparameters are optimised with an evolution strategy. The phoneme and word error rates of the resulting G2P converters are presented and discussed. The tool, the processed phonetic lexicons and trained G2P models are made freely available.

Designing a Synthesized Content Feed System for Community Radio

Authors: Adriana Stan

The use of text-to-speech to generate radio content is largely unexplored, despite the importance of radio in remote parts of the world, where TTS offers a robust means of transforming data into media for low-literate audiences and those without regular internet access. How suitable are TTS voices for meeting the expectations of radio listeners and what type of content are these voices best suited to deliver? We present an application for generating automated daily synthesized weather forecasts for selected locations and language varieties, based on the provision of a regularly updated weather data service. We present results from a pilot listener study aimed at exploring people’s reactions to this and other synthesized audio content, as we begin to explore best practices around the design of a synthesized content feed system for community radio.

Parallel implementation of a PIC simulation algorithm using OpenMP

Authors: Alin Suciu, Anca Nicoleta Mărginean, Anca Hangan

Particle-in-cell (PIC) simulations are focusing on the individual trajectories of a very large number of particles in self-consistent and external electric and magnetic fields; they are widely used in the study of plasma jets, for example.The main disadvantage of PIC simulations is the large simulation runtime, which often requires a parallel implementation of the algorithm.The current paper focuses on a PIC1d3v simulation algorithm [1][2] and describes the successful implementation of a parallel version of it on a multicore architecture, using OpenMP, with very promising experimental and theoretical results.

Commonsense Statements Identification and Explanation with Transformer-based Encoders

Authors: Anca Nicoleta Mărginean

In this work, we present our empirical attempt to identify the proper strategy of using Transformer Language Models to identify sentences consistent with commonsense.We tackle the first two tasks from the ComVE (Wang et al., 2020a) competition.The starting point for our work is the BERT assumption according to which a large number of NLP tasks can be solved with pre-trained Transformers with no substantial task-specific changes of the architecture.However, our experiments show that the encoding strategy can have a great impact on the quality of the fine-tuning.The combination between cross-encoding and multi-input models worked better than one cross-encoder and allowed us to achieve comparable results with the state-of-the-art without the use of any external data.

Smart Environments and Social Robots for Age-Friendly Integrated Care Services

Authors: Tudor Cioara, Anghel Ionuț

The world is facing major societal challenges because of an aging population that is putting increasing pressure on the sustainability of care. While demand for care and social services is steadily increasing, the supply is constrained by the decreasing workforce. The development of smart, physical, social and age-friendly environments is identified by World Health Organization (WHO) as a key intervention point for enabling older adults, enabling them to remain as much possible in their residences, delay institutionalization, and ultimately, improve quality of life. In this study, we survey smart environments, machine learning and robot assistive technologies that can offer support for the independent living of older adults and provide age-friendly care services. We describe two examples of integrated care services that are using assistive technologies in innovative ways to assess and deliver of timely interventions for polypharmacy management and for social and cognitive activity support in older adults. We describe the architectural views of these services, focusing on details about technology usage, end-user interaction flows and data models that are developed or enhanced to achieve the envisioned objective of healthier, safer, more independent and socially connected older people.

Blockchain and Demand Response: Zero-Knowledge Proofs for Energy Transactions Privacy

Authors: Tudor Cioara, Anghel Ionuț

Nowadays, the adoption of demand response programs is still lagging due to the prosumers' lack of awareness, fear of losing control and privacy of energy data, etc. Programs decentralization, by adopting promising technologies such as blockchain, may bring significant advantages in terms of transparency, openness, improved control, and increased active participation of prosumers. Nevertheless, even though in general the transparency of the public blockchain is a desirable feature in the energy domain, the prosumer energy data is sensitive and rather private, thus, a privacy-preserving solution is required. In this paper, we present a decentralized implementation of demand response programs on top of the public blockchain which deals with the privacy of the prosumer's energy data using zero-knowledge proofs and validates on the blockchain the prosumer's activity inside the program using smart contracts. Prosumer energy data is kept private, while on the blockchain it is stored a zero-knowledge proof that is generated by the prosumer itself allowing the implementation of functions to validate potential deviations from the request and settle prosumer's activity. The solution evaluation results are promising in terms of ensuring the privacy of prosumer energy data stored in the public blockchain and detecting potential data inconsistencies.

Smart Grid Management Using Blockchain: Future Scenarios and Challenges

Authors: Tudor Cioara, Anghel Ionuț

Decentralized management and coordination of energy systems are emerging trends facilitated by the uptake of the Internet of Things and Blockchain offering new opportunities for more secure, resilient, and efficient energy distribution. Even though the use of distributed ledger technology in the energy domain is promising, the development of decentralized smart grid management solutions is in the early stages. In this paper, we define a layered architecture of a blockchain-based smart grid management platform featuring energy data metering and tamper-proof registration, business enforcement via smart contracts, and Oracle-based integration of high computational services supporting the implementation of future grid management scenarios. Three such scenarios are discussed from the perspective of their implementation using the proposed blockchain platform and associated challenges: peer to peer energy trading, decentralized management, and aggregation of energy flexibility and operation of community oriented Virtual Power Plants.

Data Centers Optimized Integration with Multi-Energy Grids: Test Cases and Results in Operational Environment

Authors: Tudor Cioara, Anghel Ionuț

In this paper, we address the management of Data Centers (DCs) by considering their optimal integration with the electrical, thermal, and IT (Information Technology) networks helping them to meet sustainability objectives and gain primary energy savings. Innovative scenarios are defined for exploiting the DCs electrical, thermal, and workload flexibility as a commodity and Information and Communication Technologies (ICT) are proposed and used as enablers for the scenarios’ implementation. The technology and scenarios were evaluated in the context of two operational DCs: a micro DC in Poznan which has on-site renewable sources and a DC in Point Saint Martin. The test cases’ results validate the possibility of using renewable energy sources (RES) for exploiting DCs’ energy flexibility and the potential of combining IT load migration with the availability of RES to increase the amount of energy flexibility by finding a trade-off between the flexibility level, IT load Quality of Service (QoS), and the RES production level. Moreover, the experiments conducted show that the DCs can successfully adapt their thermal energy profile for heat re-use as well as the combined electrical and thermal energy profiles to match specific flexibility requests.

Energy Flexibility Prediction for Data Center Engagement in Demand Response Programs

Authors: Tudor Cioara, Anghel Ionuț

In this paper, we address the problem of the efficient and sustainable operation of data centers (DCs) from the perspective of their optimal integration with the local energy grid through active participation in demand response (DR) programs. For DCs’ successful participation in such programs and for minimizing the risks for their core business processes, their energy demand and potential flexibility must be accurately forecasted in advance. Therefore, in this paper, we propose an energy prediction model that uses a genetic heuristic to determine the optimal ensemble of a set of neural network prediction models to minimize the prediction error and the uncertainty concerning DR participation. The model considers short term time horizons (i.e., day-ahead and 4-h-ahead refinements) and different aspects such as the energy demand and potential energy flexibility (the latter being defined in relation with the baseline energy consumption). The obtained results, considering the hardware characteristics as well as the historical energy consumption data of a medium scale DC, show that the genetic-based heuristic improves the energy demand prediction accuracy while the intra-day prediction refinements further reduce the day-ahead prediction error. In relation to flexibility, the prediction of both above and below baseline energy flexibility curves provides good results for the mean absolute percentage error (MAPE), which is just above 6%, allowing for safe DC participation in DR programs.

Adapted Binary Particle Swarm Optimization for Efficient Features Selection in the Case of Imbalanced Sensor Data

Authors: Tudor Cioara, Anghel Ionuț

Daily living activities (DLAs) classification using data collected from wearable monitoring sensors is very challenging due to the imbalance characteristics of the monitored data. A major research challenge is to determine the best combination of features that returns the best accuracy results using minimal computational resources, when the data is heterogeneous and not fitted for classical algorithms that are designed for balanced low-dimensional datasets. This research article: (1) presents a modification of the classical version of the binary particle swarm optimization (BPSO) algorithm that introduces a particular type of particles called sensor particles, (2) describes the adaptation of this algorithm for data generated by sensors that monitor DLAs to determine the best positions and features of the monitoring sensors that lead to the best classification results, and (3) evaluates and validates the proposed approach using a machine learning methodology that integrates the modified version of the algorithm. The methodology is tested and validated on the Daily Life Activities (DaLiAc) dataset.

The Limits of Efficiency for Open- and Closed-World Query Evaluation Under Guarded TGDs

Authors: Cristina Feier

Ontology-mediated querying and querying in the presence of constraints are two key database problems where tuple-generating dependencies (TGDs) play a central role. In ontology-mediated querying, TGDs can formalize the ontology and thus derive additional facts from the given data, while in querying in the presence of constraints, they restrict the set of admissible databases. In this work, we study the limits of efficient query evaluation in the context of the above two problems, focusing on guarded and frontier-guarded TGDs and on UCQs as the actual queries. We show that a class of ontology-mediated queries (OMQs) based on guarded TGDs can be evaluated in FPT iff the OMQs in the class are equivalent to OMQs in which the actual query has bounded treewidth, up to some reasonable assumptions. For querying in the presence of constraints, we consider classes of constraint-query specifications (CQSs) that bundle a set of constraints with an actual query. We show a dichotomy result for CQSs based on guarded TGDs that parallels the one for OMQs except that, additionally, FPT coincides with PTime combined complexity. The proof is based on a novel connection between OMQ and CQS evaluation. Using a direct proof, we also show a similar dichotomy result, again up to some reasonable assumptions, for CQSs based on frontier-guarded TGDs with a bounded number of atoms in TGD heads. Our results on CQSs can be viewed as extensions of Grohe's well-known characterization of the tractable classes of CQs (without constraints). Like Grohe's characterization, all the above results assume that the arity of relation symbols is bounded by a constant. We also study the associated meta problems, i.e., whether a given OMQ or CQS is equivalent to one in which the actual query has bounded treewidth.

When is Ontology-Mediated Querying Efficient?

Authors: Cristina Feier

In ontology-mediated querying, description logic (DL) ontologies are used to enrich incomplete data with domain knowledge which results in more complete answers to queries. However, the evaluation of ontology-mediated queries (OMQs) over relational databases is computationally hard. This raises the question when OMQ evaluation is efficient, in the sense of being tractable in combined complexity or fixed-parameter tractable. We study this question for a range of ontology-mediated query languages based on several important and widely-used DLs, using unions of conjunctive queries as the actual queries. For the DL ELHI extended with the bottom concept, we provide a characterization of the classes of OMQs that are fixed-parameter tractable. For its fragment EL extended with domain and range restrictions and the bottom concept (which restricts the use of inverse roles), we provide a characterization of the classes of OMQs that are tractable in combined complexity. Both results are in terms of equivalence to OMQs of bounded tree width and rest on a reasonable assumption from parameterized complexity theory. They are similar in spirit to Grohe's seminal characterization of the tractable classes of conjunctive queries over relational databases. We further study the complexity of the meta problem of deciding whether a given OMQ is equivalent to an OMQ of bounded tree width, providing several completeness results that range from NP to 2ExpTime, depending on the DL used. We also consider the DL-Lite family of DLs, including members that admit functional roles.

Finite-Horizon Discounted Optimal Control: Stability and Performance

Authors: Lucian Bușoniu

Motivated by (approximate) dynamic programming and model predictive control problems, we analyse the stability of deterministic nonlinear discrete-time systems whose inputs minimize a discounted finite-horizon cost. We assume that the system satisfies stabilizability and detectability properties with respect to the stage cost. Then, a Lyapunov function for the closed-loop system is constructed and a uniform semiglobal stability property is ensured, where the adjustable parameters are both the discount factor and the horizon length, which corresponds to the number of iterations for dynamic programming algorithms like value iteration. Stronger stability properties such as global exponential stability are also provided by strengthening the initial assumptions. We give bounds on the discount factor and the horizon length under which stability holds and we provide conditions under which these are less conservative than the bounds of the literature for discounted infinite-horizon cost and undiscounted finite-horizon costs, respectively. In addition, we provide new relationships between the optimal value functions of the discounted, undiscounted, infinite-horizon and finite-horizon costs respectively, which are very different from those available in the approximate dynamic programming literature. These relationships rely on assumptions that are more likely to be satisfied in a control context. Finally, we investigate stability when only a near-optimal sequence of inputs for the discounted finite-horizon cost is available, covering approximate value iteration as a particular case.

Sliding mode control of a ball balancing robot

Authors: Lucian Bușoniu, Alexandru Codrean

This paper presents a sliding mode control design for a ball-balancing robot (ballbot), with associated real-time results. The sliding mode control is designed based on the linearized plant model, and is robust to matched uncertainties. The design is considerably simpler than other nonlinear control strategies presented in the literature, and the experimental results for stabilization and tracking show much better performances than those obtained with linear control (in particular, a linear quadratic regulator).

Cross Entropy Optimization of Action Modification Policies for Continuous-Valued MDPs

Authors: Lucian Bușoniu

We propose an algorithm to search for parametrized policies in continuous state and action Markov Decision Processes (MDPs). The policies are represented via a number of basis functions, and the main novelty is that each basis function corresponds to a small, discrete modification of the continuous action. In each state, the policy chooses a discrete action modification associated with a basis function having the maximum value at the current state. Empirical returns from a representative set of initial states are estimated in simulations to evaluate the policies. Instead of using slow gradient-based algorithms, we apply cross entropy method for updating the parameters. The proposed algorithm is applied to a double integrator and an inverted pendulum problem, with encouraging results.

The ClujUAV student competition: A corridor navigation challenge with autonomous drones

Authors: Lucian Bușoniu, Levente Tamás

We describe a novel student contest concept in which an unmanned aerial vehicle (UAV or drone) must autonomously navigate a straight corridor using feedback from camera images. The objective of the contest is to promote engineering skills (related to sensing and control in particular) among students and young professionals, by means of an attractive robotics topic in an exciting competition format. The first edition of this contest was organized in Cluj-Napoca, Romania on October 19th 2019. Teams from industry and academia competed, with an overall positive experience. We outline the challenge and scoring rules, together with the technical solutions of the teams, and close with a summary of the results and points to improve for the next editions.

Sorting Objects from a Conveyor Belt Using POMDPs with Multiple-Object Observations and Information-Gain Rewards

Authors: Lucian Bușoniu, Levente Tamás

We consider a robot that must sort objects transported by a conveyor belt into different classes. Multiple observations must be performed before taking a decision on the class of each object, because the imperfect sensing sometimes detects the incorrect object class. The objective is to sort the sequence of objects in a minimal number of observation and decision steps. We describe this task in the framework of partially observable Markov decision processes, and we propose a reward function that explicitly takes into account the information gain of the viewpoint selection actions applied. The DESPOT algorithm is applied to solve the problem, automatically obtaining a sequence of observation viewpoints and class decision actions. Observations are made either only for the object on the first position of the conveyor belt or for multiple adjacent positions at once. The performance of the single- and multiple-position variants is compared, and the impact of including the information gain is analyzed. Real-life experiments with a Baxter robot and an industrial conveyor belt are provided.

When to stop value iteration: stability and near-optimality versus computation

Authors: Lucian Bușoniu

Value iteration (VI) is a ubiquitous algorithm for optimal control, planning, and reinforcement learning schemes. Under the right assumptions, VI is a vital tool to generate inputs with desirable properties for the controlled system, like optimality and Lyapunov stability. As VI usually requires an infinite number of iterations to solve general nonlinear optimal control problems, a key question is when to terminate the algorithm to produce a "good" solution, with a measurable impact on optimality and stability guarantees. By carefully analysing VI under general stabilizability and detectability properties, we provide explicit and novel relationships of the stopping criterion's impact on near-optimality, stability and performance, thus allowing to tune these desirable properties against the induced computational cost. The considered class of stopping criteria encompasses those encountered in the control, dynamic programming and reinforcement learning literature and it allows considering new ones, which may be useful to further reduce the computational cost while endowing and satisfying stability and near-optimality properties. We therefore lay a foundation to endow machine learning schemes based on VI with stability and performance guarantees, while reducing computational complexity.

Learning control for transmission and navigation with a mobile robot under unknown communication rates

Authors: Lucian Bușoniu, Alexandru Codrean

In tasks such as surveying or monitoring remote regions, an autonomous robot must move while transmitting data over a wireless network with unknown, position-dependent transmission rates. For such a robot, this paper considers the problem of transmitting a data buffer in minimum time, while possibly also navigating towards a goal position. Two approaches are proposed, each consisting of a machine-learning component that estimates the rate function from samples; and of an optimal-control component that moves the robot given the current rate function estimate. Simple obstacle avoidance is performed for the case without a goal position. In extensive simulations, these methods achieve competitive performance compared to known-rate and unknown-rate baselines. A real indoor experiment is provided in which a Parrot AR.Drone 2 successfully learns to transmit the buffer.

Space-time budget allocation policy design for viral marketing

Authors: Lucian Bușoniu

We address formally the problem of opinion dynamics when the agents of a social network (e.g., consumers) are not only influenced by their neighbors but also by an external influential entity referred to as a marketer. The influential entity tries to sway the overall opinion as close as possible to a desired opinion by using a specific influence budget. We assume that the exogenous influences of the entity happen during discrete-time advertising campaigns; consequently, the overall closed-loop opinion dynamics becomes a linear-impulsive (hybrid) one. The main technical issue addressed is finding how the marketer should allocate its budget over time (through marketing campaigns) and over space (among the agents) such that the agents' opinion be as close as possible to the desired opinion. Our main results show that the marketer has to prioritize certain agents over others based on their initial condition, their influence power in the social graph and the size of the cluster they belong to. The corresponding space-time allocation problem is formulated and solved for several special cases of practical interest. Valuable insights can be extracted from our analysis. For instance, for most cases, we prove that the marketer has an interest in investing most of its budget at the beginning of the process and that budget should be shared among agents according to the famous water-filling allocation rule. Numerical examples illustrate the analysis.

Comparison of Deep-Learning and Conventional Machine-Learning Methods for the Automatic Recognition of the Hepatocellular Carcinoma Areas from Ultrasound Images

Authors: Sergiu Nedevschi, Raluca Didona Brehar, Delia Alexandrina Mitrea, Tiberiu Marița

The emergence of deep-learning methods in different computer vision tasks has proved to offer increased detection, recognition or segmentation accuracy when large annotated image datasets are available. In the case of medical image processing and computer-aided diagnosis within ultrasound images, where the amount of available annotated data is smaller, a natural question arises: are deep-learning methods better than conventional machine-learning methods? How do the conventional machine-learning methods behave in comparison with deep-learning methods on the same dataset? Based on the study of various deep-learning architectures, a lightweight multi-resolution Convolutional Neural Network (CNN) architecture is proposed. It is suitable for differentiating, within ultrasound images, between the Hepatocellular Carcinoma (HCC), respectively the cirrhotic parenchyma (PAR) on which HCC had evolved. The proposed deep-learning model is compared with other CNN architectures that have been adapted by transfer learning for the ultrasound binary classification task, but also with conventional machine-learning (ML) solutions trained on textural features. The achieved results show that the deep-learning approach overcomes classical machine-learning solutions, by providing a higher classification performance.

Stabilization and Validation of 3D Object Position Using Multimodal Sensor Fusion and Semantic Segmentation

Authors: Sergiu Nedevschi, Mircea Paul Mureșan

The stabilization and validation process of the measured position of objects is an important step for high-level perception functions and for the correct processing of sensory data. The goal of this process is to detect and handle inconsistencies between different sensor measurements, which result from the perception system. The aggregation of the detections from different sensors consists in the combination of the sensorial data in one common reference frame for each identified object, leading to the creation of a super-sensor. The result of the data aggregation may end up with errors such as false detections, misplaced object cuboids or an incorrect number of objects in the scene. The stabilization and validation process is focused on mitigating these problems. The current paper proposes four contributions for solving the stabilization and validation task, for autonomous vehicles, using the following sensors: trifocal camera, fisheye camera, long-range RADAR (Radio detection and ranging), and 4-layer and 16-layer LIDARs (Light Detection and Ranging). We propose two original data association methods used in the sensor fusion and tracking processes. The first data association algorithm is created for tracking LIDAR objects and combines multiple appearance and motion features in order to exploit the available information for road objects. The second novel data association algorithm is designed for trifocal camera objects and has the objective of finding measurement correspondences to sensor fused objects such that the super-sensor data are enriched by adding the semantic class information. The implemented trifocal object association solution uses a novel polar association scheme combined with a decision tree to find the best hypothesis-measurement correlations. Another contribution we propose for stabilizing object position and unpredictable behavior of road objects, provided by multiple types of complementary sensors, is the use of a fusion approach based on the Unscented Kalman Filter and a single-layer perceptron. The last novel contribution is related to the validation of the 3D object position, which is solved using a fuzzy logic technique combined with a semantic segmentation image. The proposed algorithms have a real-time performance, achieving a cumulative running time of 90 ms, and have been evaluated using ground truth data extracted from a high-precision GPS (global positioning system) with 2 cm accuracy, obtaining an average error of 0.8 m.

Integration of Real-Time Image Fusion in the Robotic-Assisted Treatment of Hepatocellular Carcinoma

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea, Tiberiu Marița, Doina Liana Pisla

Hepatocellular carcinoma (HCC) is one of the leading causes of cancer-related deaths worldwide, with its mortality rate correlated with the tumor staging; i.e., early detection and treatment are important factors for the survival rate of patients. This paper presents the development of a novel visualization and detection system for HCC, which is a composing module of a robotic system for the targeted treatment of HCC. The system has two modules, one for the tumor visualization that uses image fusion (IF) between computerized tomography (CT) obtained preoperatively and real-time ultrasound (US), and the second module for HCC automatic detection from CT images. Convolutional neural networks (CNN) are used for the tumor segmentation which were trained using 152 contrast-enhanced CT images. Probabilistic maps are shown as well as 3D representation of HCC within the liver tissue. The development of the visualization and detection system represents a milestone in testing the feasibility of a novel robotic system in the targeted treatment of HCC. Further optimizations are planned for the tumor visualization and detection system with the aim of introducing more relevant functions and increase its accuracy.

Adversarial Graph Learning and Deep Learning Techniques for improving diagnosis within CT and Ultrasound images

Authors: Delia Alexandrina Mitrea, Sergiu Nedevschi

Computed tomography (CT) examination has an important role in diagnosing, monitoring, and evaluating treatment response in patients infected with COVID-19. Chest CT images provide a reliable alternative to the reverse transcription polymerase chain reaction (RT-PCR). Ultrasonography provides a non-invasive and low-cost solution for diagnosing liver tumors. Recognition of hepatocellular carcinoma (HCC) in ultrasound images can provide a suitable alternative to the more invasive and dangerous method of doing a biopsy. Significant effort has been dedicated to the development of automated ways of diagnosing COVID-19 based on CT scans and on diagnosing liver tumors based on ultrasound images, some of them use deep learning methods. This paper proposes a method for training deep neural networks for performing binary classification, which is called Adversarial Graph Learning. Improved classification performance and reliability is achieved by using our method for both target tasks, recognizing COVID-19 in CT images and recognizing HCC in ultrasound images.

IEEE Transactions on Intelligent Transportation Systems publication information

Authors: Sergiu Nedevschi

Climate change opinions in online debate sites

Authors: Adrian Petru Groza, Anca Nicoleta Mărginean

Debate sites in social media provide a unified platform for citizens to discuss controversial questions and to put forward their ideas and arguments on the issues of common interest. Opinions of citizens may provide useful knowledge to stakeholders but manual analysis of arguments in debate sites is tedious, while computational support to this end has been rather scarce. We focus here on developing a technical instrumentation for making sense of a set of online arguments and aggregating them into usable results for policy making and climate science communication. Our objectives are: (i) to aggregate arguments posted for a certain debate topic, (ii) to consolidate opinions posted under several but related topics either in the same or different debate site, and (iii) to identify possible linguistic characteristics of the argumentative texts. For the first objective, we propose a voting method based on subjective logic [13]. For the second objective, we assess the semantic similarity between two debate topics based on textual entailment [28]. For the third objective, we employ various existing methods for lexical analysis such as frequency analysis or readability indexes. Although we focused here on the climate change, the method can be applied to any domain.

The OpenModelica Integrated Modeling, Simulation, and Optimization Environment

Authors: Adrian Pop

OpenModelica is currently the most complete opensource Modelica-and FMI-based modeling, simulation, optimization, and model-based development environment.Moreover, the OpenModelica environment provides a number of facilities such as debugging; optimization; visualization and 3D animation; web-based model editing and simulation; scripting from Modelica, Python, Julia, and Matlab; efficient simulation and co-simulation of FMI-based models; compilation for embedded systems; Modelica-UML integration; requirement verification; and generation of parallel code for multi-ore architectures.The environment is based on Modelica and uses an extended version of Modelica for its implementation.This overview paper intends to give an up-to-date brief description of the capabilities of the system, and the main vision behind its development.

OMSimulator - Integrated FMI and TLM-based Co-simulation with Composite Model Editing and SSP

Authors: Adrian Pop

OMSimulator is an FMI-based co-simulation tool and recent addition to the OpenModelica tool suite.It supports large-scale simulation and virtual prototyping using models from multiple sources utilizing the FMI standard.It is integrated into OpenModelica but also available stand-alone, i.e., without dependencies to Modelicaspecific models or technology.OMSimulator provides an industrial-strength open-source FMI-based modelling and simulation tool.Input/output ports of FMUs can be connected, ports can be grouped to buses, FMUs can be parameterized and composed, and composite models can be exported according to the (preliminary) SSP (System Structure and Parameterization) standard.Efficient FMIbased simulation is provided for both model-exchange and co-simulation.TLM-based tool connection is provided for a range of applications, e.g., Adams, Simulink, Beast, Dymola, and OpenModelica.Moreover, optional TLM (Transmission Line Modelling) domain-specific connectors are also supported, providing additional numerical stability to co-simulation.An external API is available for use from other tools and scripting languages such as Python and Lua.The paper gives an overview of the tool functionality, compares with related work, and presents experience from industrial usage.

A New OpenModelica Compiler High Performance Frontend

Authors: Adrian Pop

The equation-based object-oriented Modelica language allows easy composition of models from components.It is very easy to create very large parametrized models using component arrays of models.Current open-source and commercial Modelica tools can with ease handle models with a hundred thousand equations and a thousand states.However, when the system size goes above half a million (or more) equations the tools begin to have problems with scalability.This paper presents the new frontend of the OpenModelica compiler, designed with scalability in mind.The new OpenModelica frontend can handle much larger systems than the current one with better time and memory performance.The new frontend was validated against large models from the ScalableTestSuite library and Modelica Standard Library, with good results.

Controller Design for a Magnetic Levitation Kit using OpenModelica's Integration with the Julia Language

Authors: Adrian Pop

This paper presents a practical application of computer aided control systems design using a new OpenModelica API (OMJulia) which allows to conveniently operate on Modelica models from the Julia language.Julia is a rather young language (Julia 1.0 was released in August 2018) designed to address the needs of numerical analysis and computational science, in particular it already has decent support for the control community.The magnetic levitation application at hand demonstrates how control system design can benefit from a suitable integration between Julia and Modelica.It is based on a commercially available control education kit in which the original controller is replaced by our own digital controller developed in this work.There exists an accompanying but independent paper which introduces the complete OMJulia API.

OMJulia: An OpenModelica API for Julia-Modelica Interaction

Authors: Adrian Pop

Modelica is an object oriented, acausal equation-based language for describing complex, hybrid dynamic models.About ten Modelica implementations exist, of which most are commercial and two are open source; the implementations have varying levels of tool functionality.Many Modelica implementations have limited support for model analysis.It is therefore of interest to integrate Modelica tools with a powerful scripting and programming language, such as Julia.Julia is a modern and free language for scientific computing.Such integration would facilitate the needed analysis possibilities and can speed up the development of effient simulation models.A number of design choices for interaction between Julia and Modelica tools are discussed.Next, Julia package OMJulia is introduced with an API for interaction between Open-Modelica and Julia.Some discussion of the reasoning behind the OMJulia design is given.The API is based on a new class ModelicaSystem within package OMJulia, with systematic methods which operate on instantiated models.OMJulia supports handling of FMU and Modelica models, setting and getting model values, as well as some model operations.Results are available in Julia for further analysis.OMJulia is a further development of a previous OMPython package; a key advantage of Julia over Python is that Julia has better support for control engineering packages.OMJulia represents a first effort to interface a relatively complete Modelica tool to Julia, giving access to an open source set-up for modeling and analysis, including control synthesis, easily installable from a unified package manager.Some possibilities of OMJulia are illustrated by application to a few simple, yet industrially relevant problems within control design.

MetaModelica - A Symbolic-Numeric Modelica Language and Comparison to Julia

Authors: Adrian Pop

The need for integrating system modeling with advanced tool capabilities is becoming increasingly pronounced.For example, a set of simulation experiments may give rise to new data that are used to systematically construct a series of new models, e.g. for further simulation and design optimization.Such combined symbolic-numeric capabilities have been pioneered by dynamically typed interpreted languages such as Lisp and Mathematica.Such capabilities are also relevant for advanced modeling and simulation applications but lacking in the standard Modelica language.Therefore, this is a topic of long-running design discussions in the Modelica Design group.One contribution in this direction is MetaModelica, that has been developed to extend Modelica with symbolic operations and advanced data structures, while preserving safe engineering practices through static type checking and a compilation-based efficient implementation.Another recent effort is Modia, implemented using the Julia macro mechanism, making it dynamically typed but also adding new capabilities.The Julia language has appeared rather recently and has expanded into a large and fast-growing ecosystem.It is dynamically typed, provides both symbolic and numeric operations, advanced data structures, and has a just-intime compilation-based efficient implementation.Despite independent developments there are surprisingly many similarities between Julia and MetaModelica.This paper presents MetaModelica and its environment as a large case study, together with a short comparison to Julia.Since Julia may be important for the future Modelica, some integration options between Modelica tools and Julia are also discussed, including a possible approach for implementing MetaModelica (and OpenModelica) in Julia.

Modelica On The Web

Authors: Adrian Pop

Modelica has been around as a language from the late 1990's and since then a range of compilers and editors have emerged.Currently none of these environments provide a web-based user interface and follow the approach of requiring each end-user to install the application (typically together with a set of dependencies) on their local machine.This in itself may or may not be of major concern.Of more importance is their current lack of a seamless collaborative approach to modeling.This paper presents the first web-based collaborative graphical and textual modeling environment for Modelica based on WebGME and OpenModelica.Graphical composition of Modelica models from component libraries is supported via WebGME.Textual editing of the composite model is possible via OMWebBook.

Blockchain-Based Scalable and Tamper-Evident Solution for Registering Energy Data

Authors: Tudor Cioara, Anghel Ionuț

Nowadays, it has been recognized that blockchain can provide the technological infrastructure for developing decentralized, secure, and reliable smart energy grid management systems. However, an open issue that slows the adoption of blockchain technology in the energy sector is the low scalability and high processing overhead when dealing with the real-time energy data collected by smart energy meters. Thus, in this paper, we propose a scalable second tier solution which combines the blockchain ledger with distributed queuing systems and NoSQL (Not Only SQL database) databases to allow the registration of energy transactions less frequently on the chain without losing the tamper-evident benefits brought by the blockchain technology. At the same time, we propose a technique for tamper-evident registration of smart meters' energy data and associated energy transactions using digital fingerprinting which allows the energy transaction to be linked hashed-back on-chain, while the sensors data is stored off-chain. A prototype was implemented using Ethereum and smart contracts for the on-chain components while for the off-chain components we used Cassandra database and RabbitMQ messaging broker. The prototype proved to be effective in managing a settlement of energy imbalances use-case and during the evaluation conducted in simulated environment shows promising results in terms of scalability, throughput, and tampering of energy data sampled by smart energy meters.

A Fog Computing Enabled Virtual Power Plant Model for Delivery of Frequency Restoration Reserve Services

Authors: Tudor Cioara, Anghel Ionuț

Nowadays, centralized energy grid systems are transitioning towards more decentralized systems driven by the need for efficient local integration of new deployed small scale renewable energy sources. The high limits for accessing the energy markets and also for the delivery of ancillary services act as a barrier for small scale prosumers participation forcing the implementation of new cooperative business models at the local level. This paper is proposing a fog computing infrastructure for the local management of energy systems and the creation of coalitions of prosumers able to provide ancillary services to the grid. It features an edge devices layer for energy monitoring of individual prosumers, a fog layer providing Information and Communication Technologies (ICT) techniques for managing local energy systems by implementing cooperative models, and a cloud layer where the service specific technical requirements are defined. On top, a model has been defined allowing the dynamical construction of coalitions of prosumers as Virtual Power Plants at the fog layer for the provisioning of frequency restoration reserve services while considering both the prosumers' local constraints and the service ones as well as the constituents' profit maximization. Simulation results show our solution effectiveness in selecting the optimal coalition of prosumers to reliably deliver the service meeting the technical constraints while featuring a low time and computation overhead being feasible to be run closer to the edge.

Reuse of Data Center Waste Heat in Nearby Neighborhoods: A Neural Networks-Based Prediction Model

Authors: Tudor Cioara, Anghel Ionuț

This paper addresses the problem of data centers’ cost efficiency considering the potential of reusing the generated heat in district heating networks. We started by analyzing the requirements and heat reuse potential of a high performance computing data center and then we had defined a heat reuse model which simulates the thermodynamic processes from the server room. This allows estimating by means of Computational Fluid Dynamics simulations the temperature of the hot air recovered by the heat pumps from the server room allowing them to operate more efficiently. To address the time and space complexity at run-time we have defined a Multi-Layer Perceptron neural network infrastructure to predict the hot air temperature distribution in the server room from the training data generated by means of simulations. For testing purposes, we have modeled a virtual server room having a volume of 48 m3 and two typical 42U racks. The results show that using our model the heat distribution in the server room can be predicted with an error less than 1 °C allowing data centers to accurately estimate in advance the amount of waste heat to be reused and the efficiency of heat pump operation.

Rewritability in Monadic Disjunctive Datalog, MMSNP, and Expressive Description Logics

Authors: Cristina Feier

We study rewritability of monadic disjunctive Datalog programs, (the complements of) MMSNP sentences, and ontology-mediated queries (OMQs) based on expressive description logics of the ALC family and on conjunctive queries. We show that rewritability into FO and into monadic Datalog (MDLog) are decidable, and that rewritability into Datalog is decidable when the original query satisfies a certain condition related to equality. We establish 2NExpTime-completeness for all studied problems except rewritability into MDLog for which there remains a gap between 2NExpTime and 3ExpTime. We also analyze the shape of rewritings, which in the MMSNP case correspond to obstructions, and give a new construction of canonical Datalog programs that is more elementary than existing ones and also applies to formulas with free variables.

Stability guarantees for nonlinear discrete-time systems controlled by approximate value iteration

Authors: Lucian Bușoniu

Value iteration is a method to generate optimal control inputs for generic nonlinear systems and cost functions. Its implementation typically leads to approximation errors, which may have a major impact on the closed-loop system performance. We talk in this case of approximate value iteration (AVI). In this paper, we investigate the stability of systems for which the inputs are obtained by AVI. We consider deterministic discrete-time nonlinear plants and a class of general, possibly discounted, costs. We model the closed-loop system as a family of systems parameterized by tunable parameters, which are used for the approximation of the value function at different iterations, the discount factor and the iteration step at which we stop running the algorithm. It is shown, under natural stabilizability and detectability properties as well as mild conditions on the approximation errors, that the family of closed-loop systems exhibit local practical stability properties. The analysis is based on the construction of a Lyapunov function given by the sum of the approximate value function and the Lyapunov-like function that characterizes the detectability of the system. By strengthening our conditions, asymptotic and exponential stability properties are guaranteed.

Learning-based control for a communicating mobile robot under unknown rates

Authors: Lucian Bușoniu

In problems such as surveying or monitoring remote regions, a mobile robot must transmit data over a wireless network with unknown, position-dependent transmission rates. We propose an algorithm to achieve this objective that learns approximations of the rate function and of an optimal-control solution that transmits the data in minimum time. The rates are estimated with supervised learning from the samples observed; and the control is found with dynamic programming sweeps around the current state of the robot that exploit the rate function estimate, combined with online reinforcement learning. For both synthetic and realistic rate functions, our experiments show that the learning algorithm empties the data buffer in less than twice the number of steps achieved by a model-based solution that requires to perfectly know the rate function.

Optimistic minimax search for noncooperative switched control with or without dwell time

Authors: Lucian Bușoniu

Robust Observer-Based Tracking Control Design for Power-Assisted Wheelchairs

Authors: Lucian Bușoniu

Power-assisted wheelchairs (PAW) are efficient means of transportation for disabled persons. The resulting human-machine system includes several unknown parameters such as the mass of the user or ground adhesion. Moreover, the torque signals produced by the human are required to design a robust assistive strategy, but measuring them with torque sensors increases significantly the cost of the system. Therefore, we propose a robust observer-based assistive controller using a polytopic representation. The closed-loop control design is composed of two elements: a state feedback controller with the full state at the input, and an unknown input observer to estimate human torques and feed them into the obtained controller. The goal is to guarantee an imposed H ∞ estimation performance while achieving reference tracking. To achieve the predefined performance, the observer gains are computed by solving an LMI problem. Finally, simulation results validate the control design. The methodology follows patent WO2015173094 issued in 2015 (Mohammad et al. 2015).

Optimistic planning for the near-optimal control of nonlinear switched discrete-time systems with stability guarantees

Authors: Lucian Bușoniu

Originating in the artificial intelligence literature, optimistic planning (OP) is an algorithm that generates near-optimal control inputs for generic nonlinear discrete-time systems whose input set is finite. This technique is therefore relevant for the near-optimal control of nonlinear switched systems, for which the switching signal is the control. However, OP exhibits several limitations, which prevent its application in a standard control context. First, it requires the stage cost to take values in [0, 1], an unnatural prerequisite as it excludes, for instance, quadratic stage costs. Second, it requires the cost function to be discounted. Third, it applies for reward maximization, and not cost minimization. In this paper, we modify OP to overcome these limitations, and we call the new algorithm OPmin. We then make stabilizability and detectability assumptions, under which we derive near-optimality guarantees for OPmin and we show that the obtained bound has major advantages compared to the bound originally given by OP. In addition, we prove that a system whose inputs are generated by OPmin in a receding-horizon fashion exhibits stability properties. As a result, OPmin provides a new tool for the near-optimal, stable control of nonlinear switched discrete-time systems for generic cost functions.

Near-optimal control of nonlinear systems with simultaneous controlled and random switches

Authors: Lucian Bușoniu

We consider dual switched systems, in which two switching signals act simultaneously to select the dynamical mode. The first signal is controlled and the second is random, with probabilities that evolve either periodically or as a function of the dwell time. We formalize both cases as Markov decision processes, which allows them to be solved with a simple approximate dynamic programming algorithm. We illustrate the framework in a problem where the random signal is a delay on the control channel that is used to send the controlled signal to the system.

Forecasting Cryptocurrency Value by Sentiment Analysis: An HPC-Oriented Survey of the State-of-the-Art in the Cloud Era

Authors: Tudor Cioara

This chapter surveys the state-of-the-art in forecasting cryptocurrency value by Sentiment Analysis. Key compounding perspectives of current challenges are addressed, including blockchains, data collection, annotation, and filtering, and sentiment analysis metrics using data streams and cloud platforms. We have explored the domain based on this problem-solving metric perspective, i.e., as technical analysis, forecasting, and estimation using a standardized ledger-based technology. The envisioned tools based on forecasting are then suggested, i.e., ranking Initial Coin Offering (ICO) values for incoming cryptocurrencies, trading strategies employing the new Sentiment Analysis metrics, and risk aversion in cryptocurrencies trading through a multi-objective portfolio selection. Our perspective is rationalized on the perspective on elastic demand of computational resources for cloud infrastructures.

Table of contents

Authors: Sergiu Nedevschi

API for Accessing OpenModelica Models from Python

Authors: Adrian Pop

This paper describes a new API for operating on Modelica models in Python, through OpenModelica.Modelica is an object oriented, acausal language for describing dynamic models in the form of Differential Algebraic Equations.Modelica and various implementations such as OpenModelica have limited support for model analysis, and it is of interest to integrate Modelica code with scripting languages such as Python, which facilitate the needed analysis possibilities.The API is based on a new class Mod-elicaSystem within package OMPython of OpenModelica, with methods that operate on instantiated models.Emphasis has been put on specification of a systematic structure for the various methods of the class.A simple case study involving a water tank is used to illustrate the basic ideas.

Parallel Simulation of PDE-based Modelica Models using ParModelica

Authors: Adrian Pop

The Modelica language is a modelling and programming language for modelling cyber-physical systems using equations and algorithms.In this thesis two suggested extensions of the Modelica language are covered.Those are Partial Differential Equations (PDE) and explicit parallelism in algorithmic code.While PDEs are not yet supported by the Modelica language, this article presents a framework for solving PDEs using the algorithmic part of the Modelica language, including parallel extensions.Different numerical solvers have been implemented using the explicit parallel constructs suggested for Modelica by the ParModelica language extensions, and implemented as part of OpenModelica.The solvers have been evaluated using different models, and it can be seen how bigger models are suitable for a parallel solver.The intention has been to write a framework suitable for modelling and parallel simulation of PDEs.This work can, however, also be seen as a case study of how to write a custom solver using parallel algorithmic Modelica and how to evaluate the performance of a parallel solver.

Blockchain Based Decentralized Management of Demand Response Programs in Smart Energy Grids

Authors: Tudor Cioara, Anghel Ionuț

In this paper, we investigate the use of decentralized blockchain mechanisms for delivering transparent, secure, reliable, and timely energy flexibility, under the form of adaptation of energy demand profiles of Distributed Energy Prosumers, to all the stakeholders involved in the flexibility markets (Distribution System Operators primarily, retailers, aggregators, etc.). In our approach, a blockchain based distributed ledger stores in a tamper proof manner the energy prosumption information collected from Internet of Things smart metering devices, while self-enforcing smart contracts programmatically define the expected energy flexibility at the level of each prosumer, the associated rewards or penalties, and the rules for balancing the energy demand with the energy production at grid level. Consensus based validation will be used for demand response programs validation and to activate the appropriate financial settlement for the flexibility providers. The approach was validated using a prototype implemented in an Ethereum platform using energy consumption and production traces of several buildings from literature data sets. The results show that our blockchain based distributed demand side management can be used for matching energy demand and production at smart grid level, the demand response signal being followed with high accuracy, while the amount of energy flexibility needed for convergence is reduced.

Transforming Data Centers in Active Thermal Energy Players in Nearby Neighborhoods

Authors: Tudor Cioara, Anghel Ionuț

In this paper, we see the Data Centers (DCs) as producers of waste heat integrated with smart energy infrastructures, heat which can be re-used for nearby neighborhoods. We provide a model of the thermo-electric processes within DCs equipped with heat reuse technology, allowing them to adapt their thermal response profile to meet various levels of hot water demand. On top of the model, we have implemented computational fluid dynamics-based simulations to determine the cooling system operational parameters settings, which allow the heat to build up without endangering the servers’ safety operation as well as the distribution of the workload on the servers to avoid hot spots. This will allow for setting higher temperature set points for short periods of time and using pre-cooling and post-cooling as flexibility mechanisms for DC thermal profile adaptation. To reduce the computational time complexity, we have used neural networks, which are trained using the simulation results. Experiments have been conducted considering a small operational DC featuring a server room of 24 square meters and 60 servers organized in four racks. The results show the DCs’ potential to meet different levels of thermal energy demand by re-using their waste heat in nearby neighborhoods.

Performance Comparison of Infrared and Ultrasonic Sensors in Distance Measuring

Authors: Anghel Ionuț

Ultrasonic and infrared sensors are widely used for contact-less mid-range distance measurements, in navigation systems, robotics and vehicle related applications.This paper showcases the performance analysis of infrared and ultrasonic sensors in distance measuring for obstacles made of different types of materials such as: sponge, cardboard, wood, plastic, metal and glass, materials with different textures, thickness, colors and reflexion. Navigation systems are using both sensors for locating as well as for avoiding different types of obstacles.Aspects of weight, dimensions, costs, accuracy, response times, and others make sensors widely accepted in all areas. It is extremely important to choose the best sensor specific to a particular measuring environment but at the same time that sensor that can calculate the distance for different types of obstacles. The analysis carried out during this paper aims to provide guidance in choosing the appropriate sensor for a certain type of obstacle.

From Conjunctive Queries to Instance Queries in Ontology-Mediated Querying

Authors: Cristina Feier

We consider ontology-mediated queries (OMQs) based on expressive description logics of the ALC family and (unions) of conjunctive queries, studying the rewritability into OMQs based on instance queries (IQs). Our results include exact characterizations of when such a rewriting is possible and tight complexity bounds for deciding rewritability. We also give a tight complexity bound for the related problem of deciding whether a given MMSNP sentence (in other words: the complement of a monadic disjunctive Datalog program) is equivalent to a constraint satisfaction problem.

Reinforcement learning for control: Performance, stability, and deep approximators

Authors: Lucian Bușoniu

Fall monitoring and detection for at-risk persons using a UAV

Authors: Lucian Bușoniu

We describe a demonstrator application that uses a UAV to monitor and detect falls of an at-risk person. The position and state (upright or fallen) of the person are determined with deep-learning-based computer vision, where existing network weights are used for position detection, while for fall detection the last layer is fine-tuned in additional training. A simple visual servoing control strategy keeps the person in view of the drone, and maintains the drone at a set distance from the person. In experiments, falls were reliably detected, and the algorithm was able to successfully track the person indoors.

Continuous-action planning for discounted infinite-horizon nonlinear optimal control with Lipschitz values

Authors: Lucian Bușoniu

Observer-Based Assistive Control Design Under Time-Varying Sampling for Power-Assisted Wheelchairs

Authors: Lucian Bușoniu

Compared to manual wheelchairs and fully electric powered wheelchairs, power-assisted wheelchairs (PAWs) provide a special structure where the human can use her/his propulsion to interact with the assistive system. In this context, different studies have focused on the assistive control of PAWs in recent years. This paper presents an observed-based assistive control design using only position encoders. With a time-varying sampling induced by these position encoders, the wheelchair is described by a discrete-time Linear Parameter Varying model. Based on a Takagi-Sugeno (TS) representation, an observer is designed by using LMI techniques. According to the estimated human torques, we use the frequencies with which the wheels are pushed to compute the reference velocity of the centre of gravity. The wheelchair turns with a constant yaw velocity when one of two wheels is braked by the human. Reference tracking is accomplished by a PI controller. Simulation results confirm that the proposed assistive control algorithm provides a good maneuverability for users to control the velocity of the centre of gravity and the yaw velocity of the wheelchair.

Stability analysis of discrete-time finite-horizon discounted optimal control

Authors: Lucian Bușoniu

Discounted costs are considered in many fields, like reinforcement learning, for which various algorithms can be used to obtain optimal inputs for finite horizons. The related literature mostly concentrates on optimality and largely ignores stability. In this context, we study stability of general nonlinear discrete- time systems controlled by an optimal sequence of inputs that minimizes a finite-horizon discounted cost computed in a receding horizon fashion. Assumptions are made related to the stabilizability of the system and its detectability with respect to the stage cost. Then, a Lyapunov function for the closed-loop system with the receding horizon controller is constructed and a uniform semiglobal stability property is ensured, where the adjustable parameters are both the discount factor and the horizon length. Uniform global exponential stability is guaranteed by strengthening the initial assumptions, in which case explicit bounds on the discount factor and the horizon length are provided. We compare the obtained bounds in the particular cases where there is no discount or the horizon is infinite, respectively, with related results in the literature and we show our bounds improve existing ones on the examples considered.

Space-time budget allocation for marketing over social networks

Authors: Lucian Bușoniu

We address formally the problem of opinion dynamics when the agents of a social network (e.g., consumers) are not only influenced by their neighbors but also by an external influential entity referred to as a marketer. The influential entity tries to sway the overall opinion to its own side by using a specific influence budget during discrete-time advertising campaigns; consequently, the overall closed-loop dynamics becomes a linear-impulsive (hybrid) one. The main technical issue addressed is finding how the marketer should allocate its budget over time (through marketing campaigns) and over space (among the agents) such that the agents’ opinion be as close as possible to a desired opinion; for instance, the marketer may prioritize certain agents over others based on their influence in the social graph. The corresponding space-time allocation problem is formulated and solved for several special cases of practical interest. Valuable insights can be extracted from our analysis. For instance, for most cases we prove that the marketer has an interest in investing most of its budget at the beginning of the process and that budget should be shared among agents according to the famous water-filling allocation rule. Numerical examples illustrate the analysis.

Automatic Recognition of the Hepatocellular Carcinoma from Ultrasound Images using Complex Textural Microstructure Co-Occurrence Matrices (CTMCM)

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea

Italian graduates’ employability in times of economic crisis : overview, problems and possible solutions

Authors: Adriana Stan

Relying on various sources of data, the article aims to prove an overview on Italian graduates’ employment and labour market participation in times of economic crisis. It provides a short des­cription of Italy’s higher education, economy and labour market, and an analysis of graduate em­ployment, especially focusing on fresh graduates. Three possible views on the Italian case are also suggested, and some recent policy initiatives fostering graduate employability both at the national and the local levels are illustrated. In order to offer a better understanding of the topic, the Italian situation is occasionally compared to the situation of Germany, the leading economy in Europe.

Analysing debates on climate change with textual entailment and ontologies

Authors: Adrian Petru Groza

The difficult task of recognising textual entailment aims to check if a natural language text T entails a smaller statement H. Current methods rely on machine learning and various lexical resources. Our aim is to include domain knowledge when searching for entailment or non-entailment. As most available knowledge comes in form of ontologies, we focused on translating description logic axioms into lexical rules suitable for existing textual entailment algorithms. We apply the developed system in the climate change domain, where many pro and counter arguments do exist. The performed experiments indicate an increasing of performance when including domain knowledge into the existing textual entailment algorithms.

Traceability Support in OpenModelica Using Open Services for Lifecycle Collaboration (OSLC)

Authors: Adrian Pop

A common situation in industry is that a system model is composed of several sub-models which may have been developed using different tools.The quality and effectiveness of large scale system modeling heavily depends on the underlying tools used for different phases of the development lifecycle.Available modeling and simulation tools support different operations on models, such as model creation, model simulation, FMU export, model checking, and code generation.Seamless tracing of the requirements and associating them with the models and the simulation results in the context of different modeling tools is becoming increasingly important.This can be used to support several activities such as impact analysis, component reuse, verification, and validation.However, due to the lack of interoperability between tools it is often difficult to use such tools in combination.Recently, the OSLC specification has emerged for integrating different lifecycle tools using linked data.In this paper we present new work on traceability support in OpenModelica where the traceability information is exchanged with other lifecycle tools through a standardized interface and format using OSLC.In particular, OpenModelica supports automatic recording and tracing of modeling activities such as creation, modification, and destruction of models, import model description XML, export of FMUs, and creation of simulation results.

Development of a Thermodynamic Engine in OpenModelica

Authors: Adrian Pop

OpenModelica, an open source equation oriented modeling environment for steady state and dynamic simulation, lacks good chemical engineering support.This problem is addressed by making available in different ways the thermodynamic library Chemsep that comes with DWSIM, an open source sequential modular steady state simulator.Only slow speeds could be achieved through a Python-C API based interface connecting OpenModelica with the thermodynamic library.A socket programming based interface helps achieve faster speeds.Best results have been achieved by porting the thermodynamic library and the calculation routines to OpenModelica, due to two reasons: (1) thermodynamic equations are solved simultaneously with mass and energy balances (2) overheads in calling the external routines of DWSIM are eliminated.Performances of the above mentioned three approaches have been validated with steady state and dynamic simulations.Benzene -toluene separation, methanol -ethanol -water distillation, and steam distillation of an n-octane -ndecane mixture, have been carried out through these simulations.This work makes available a powerful simulation platform to the chemical engineering.

Large Scale Training through Spoken Tutorials to Promote and use OpenModelica

Authors: Adrian Pop

The step-by-step self-teaching approach through audiovideo tutorials, known as Spoken Tutorials, has been very successful.About 3.4 million students in India have taken at least one course during the past 6-year period, of which 1.6 million students have attended the rapidly expanding course programme during 2016.This programme has now been expanded by a newly developed course in Modeling and Simulation with Modelica using the OpenModelica open source tool, primarily via the OMEdit graphical user interface.The spoken tutorial programme is exclusively based on free and open source software.This paper gives an introduction to the spoken tutorial approach and presents the recently developed spoken tutorial series for Modelica using OpenModelica.Feedback of participants shows that this series is an effective tool for self-learning of OpenModelica.The paper also presents a new web version that generalises the interactive DrModelica course material, OMWebbook: it enables students to learn Modelica, do text-based modeling exercises, and run simulations without needing to install a Modelica tool.OMWebbook is also planned to be covered in a future update to the spoken tutorial course on Modelica.

An Open-Source Framework for Efficient Co-simulation of Fluid Power Systems

Authors: Adrian Pop

Simulation of fluid power systems typically requires models from multiple disciplines.Achieving accurate load dynamics for a system with complex geometry, for example, may require both a 1D model of the hydraulic circuit and a 3D multi-body model.However, most simulation tools are limited to a single discipline.A solution to these kinds of problems is co-simulation, where different tools are coupled and simulated together.Co-simulation can provide increased accuracy, improved modularity and facilitated collaboration between different organizations.Unfortunately, tool coupling typically requires tedious and error-prone manual work.It may also introduce numerical problems.For these reasons, co-simulation is often avoided as long as possible.These problems have been addressed by the development of an open-source framework for asynchronous co-simulation.Simulation tools can be interconnected through a stand-alone master simulation tool.An extensive range of tools is also supported via the Functional Mockup Interface standard.A graphical user interface has been implemented in the OpenModelica Connection Editor.System models can be created and edited from both a schematic view and a 3D view.Numerical robustness is enforced by the use of transmission line modelling.A minimalistic programming interface consisting of only two functions is used.An example model consisting of a hydraulic crane with two arms, two actuators and a hanging load is used to verify the framework.The composite model consists of nine multi-body models, one hydraulic system model and a controller.It is shown that models from various simulation tools can be replaced with a minimal amount of user input.

INTO-CPS tool chain user manual

Authors: Adrian Pop

Expert system for nutrition care process of older adults

Authors: Tudor Cioara, Anghel Ionuț

Rewritability in Monadic Disjunctive Datalog, MMSNP, and Expressive Description Logics (Invited Talk)

Authors: Cristina Feier

We study rewritability of monadic disjunctive Datalog programs, (the complements of) MMSNP sentences, and ontology-mediated queries (OMQs) based on expressive description logics of the ALC family and on conjunctive queries. We show that rewritability into FO and into monadic Datalog (MDLog) are decidable, and that rewritability into Datalog is decidable when the original query satisfies a certain condition related to equality. We establish 2NExpTime-completeness for all studied problems except rewritability into MDLog for which there remains a gap between 2NExpTime and 3ExpTime. We also analyze the shape of rewritings, which in the MMSNP case correspond to obstructions, and give a new construction of canonical Datalog programs that is more elementary than existing ones and also applies to non-Boolean queries.

Planning for optimal control and performance certification in nonlinear systems with controlled or uncontrolled switches

Authors: Lucian Bușoniu

Decoupled-dynamics distributed control for strings of nonlinear autonomous agents

Authors: Lucian Bușoniu

We introduce a novel distributed control architecture for a class of nonlinear dynamical agents moving in the “string” formation, while guaranteeing trajectory tracking and collision avoidance. Each autonomous agent uses information and relative measurements only with respect to its predecessor in the string. The performance of the scheme is entirely scalable with respect to the number of agents in formation. The scalability is a consequence of the “decoupling” of a certain bounded approximation of the closed-loop equations, entailing that individual, local analyses of the closed-loops stability at each agent will in turn guarantee the aggregated stability of the entire formation. An efficient, practical method for compensating communications induced delays is also presented.

Near-optimal control with adaptive receding horizon for discrete-time piecewise affine systems

Authors: Lucian Bușoniu

We consider the infinite-horizon optimal control of discrete-time, Lipschitz continuous piecewise affine systems with a single input. Stage costs are discounted, bounded, and use a 1 or ∞-norm. Rather than using the usual fixed-horizon approach from model-predictive control, we tailor an adaptive-horizon method called optimistic planning for continuous actions (OPC) to solve the piecewise affine control problem in receding horizon. The main advantage is the ability to solve problems requiring arbitrarily long horizons. Furthermore, we introduce a novel extension that provides guarantees on the closed-loop performance, by reusing data (“learning”) across different steps. This extension is general and works for a large class of nonlinear dynamics. In experiments with piecewise affine systems, OPC improves performance compared to a fixed-horizon approach, while the data-reuse approach yields further improvements.

Near-optimal control of nonlinear switched systems with non-cooperative switching rules

Authors: Lucian Bușoniu

International audience

The PAN-Robots Project: Advanced Automated Guided Vehicle Systems for Industrial Logistics

Authors: Sergiu Nedevschi

In modern manufacturing plants, automation is widely adopted in the production phases, which leads to a high level of productivity and efficiency. However, the same level of automation is generally not achieved in logistics, typically performed by human operators and manually driven vehicles. In fact, even though automated guided vehicles (AGVs) have been used for a few decades for goods transportation in industrial environments [1], they do not yet represent a widespread solution and are typically applied only in specific scenarios.

A Multi Patch Warping Approach for Improved Stereo Block Matching

Authors: Sergiu Nedevschi, Mircea Paul Mureșan, Radu Gabriel Dănescu

Systems of nonlinear algebraic equations with positive solutions

Authors: Sergiu Nedevschi

We are concerned with the positive solutions of an algebraic system depending on a parameter [Formula: see text] and arising in economics. For [Formula: see text] we prove that the system has at least a solution. For [Formula: see text] we give three proofs of the existence and a proof of the uniqueness of the solution. Brouwer's theorem and inequalities involving convex functions are essential tools in our proofs.

Mining arguments from cancer documents using Natural Language Processing and ontologies

Authors: Adrian Petru Groza

In the medical domain, the continuous stream of scientific research contains contradictory results supported by arguments and counter-arguments. As medical expertise occurs at different levels, part of the human agents have difficulties to face the huge amount of studies, but also to understand the reasons and pieces of evidences claimed by the proponents and the opponents of the debated topic. To better understand the supporting arguments for new findings related to current state of the art in the medical domain we need tools able to identify arguments in scientific papers. Our work here aims to fill the above technological gap. We rely on the well-known interleaving of domain knowledge with natural language processing. To formalise the existing medical knowledge, we rely on ontologies. To structure the argumentation model we use also the expressivity and reasoning capabilities of Description Logics. To perform argumentation mining we formalise various linguistic patterns in a rule-based language. We tested our solution against a corpus of scientific papers related to breast cancer. The run experiments show a F-measure between 0.71 and 0.86 for identifying conclusions of an argument and between 0.65 and 0.86 for identifying premises of an argument.

Assisting Drivers During Overtaking Using Car-2-Car Communication and Multi-Agent Systems

Authors: Adrian Petru Groza

A warning system for assisting drivers during overtaking maneuvers is proposed. The system relies on Car-2-Car communication technologies and multi-agent systems. A protocol for safety overtaking is proposed based on ACL communicative acts. The mathematical model for safety overtaking used Kalman filter to minimize localization error.

Harmonization of conflicting medical opinions using argumentation protocols and textual entailment - a case study on Parkinson disease

Authors: Adrian Petru Groza

Parkinson's disease is the second most common neurodegenerative disease, affecting more than 1.2 million people in Europe. Medications are available for the management of its symptoms, but the exact cause of the disease is unknown and there is currently no cure on the market. To better understand the relations between new findings and current medical knowledge, we need tools able to analyze published medical papers based on natural language processing and tools capable of identifying various relationships of new findings with the current medical knowledge. Our work aims to fill the above technological gap. To identify conflicting information in medical documents, we enact textual entailment technology. To encapsulate existing medical knowledge, we rely on ontologies. To connect the formal axioms in ontologies with natural text in medical articles, we exploit ontology verbalization techniques. To assess the level of disagreement between human agents with respect to a medical issue, we rely on fuzzy aggregation. To harmonize this disagreement, we design mediation protocols within a multi-agent framework.

Industrial Evaluation of Integrated Performance Analysis and Equation Model Debugging for Equation-Based Models

Authors: Adrian Pop

The ease of use and the high abstraction level of equation-based object-oriented (EOO) languages such as Modelica has the drawback that performance problems and modeling errors are often hard to nd.To address this problem, we have earlier developed advanced performance analysis and equation model debugging support in the OpenModelica tool.The aim of the work reported in this paper is to perform an independent investigation and evaluation of this equation model performance analysis and debugging methods and tool support on industrial models.The results turned out to be mainly positive.The integrated debugger and performance analyzer locates several kinds of errors such as division by zero, chattering, etc., and greatly facilitates nding the equations that take most of the execution time during simulation.It remains to further evaluate the performance proler and debugger on even larger industrial models.

Blind speech segmentation using spectrogram image-based features and Mel cepstral coefficients

Authors: Adriana Stan, Bogdan Orza, Mircea Giurgiu

This paper introduces a novel method for blind speech segmentation at a phone level based on image processing. We consider the spectrogram of the waveform of an utterance as an image and hypothesize that its striping defects, i.e. discontinuities, appear due to phone boundaries. Using a simple image destriping algorithm these discontinuities are found. To discover phone transitions which are not as salient in the image, we compute spectral changes derived from the time evolution of Mel cepstral parametrisation of speech. These socalled image-based and acoustic features are then combined to form a mixed probability function, whose values indicate the likelihood of a phone boundary being located at the corresponding time frame. The method is completely unsupervised and achieves an accuracy of 75.59% at a -3.26% over-segmentation rate, yielding an F-measure of 0.76 and an 0.80 R-value on the TIMIT dataset.

A service-based system for malnutrition prevention and self-management

Authors: Tudor Cioara, Anghel Ionuț

Stability Analysis of Discrete-Time Infinite-Horizon Optimal Control With Discounted Cost

Authors: Lucian Bușoniu

We analyze the stability of general nonlinear discrete-time systems controlled by an optimal sequence of inputs that minimizes an infinite-horizon discounted cost. First, assumptions related to the controllability of the system and its detectability with respect to the stage cost are made. Uniform semiglobal and practical stability of the closed-loop system is then established, where the adjustable parameter is the discount factor. Stronger stability properties are thereupon guaranteed by gradually strengthening the assumptions. Next, we show that the Lyapunov function used to prove stability is continuous under additional conditions, implying that stability has a certain amount of nominal robustness. The presented approach is flexible and we show that robust stability can still be guaranteed when the sequence of inputs applied to the system is no longer optimal but near-optimal. We also analyze stability for cost functions in which the importance of the stage cost increases with time, opposite to discounting. Finally, we exploit stability to derive new relationships between the optimal value functions of the discounted and undiscounted problems, when the latter is well-defined.

Model predictive control for continuous piecewise affine systems using optimistic optimization

Authors: Lucian Bușoniu

This paper considers model predictive control for continuous piecewise affine (PWA) systems. In general, this leads to a nonlinear, nonconvex optimization problem. We introduce an approach based on optimistic optimization to solve the resulting optimization problem. Optimistic optimization is based on recursive partitioning of the feasible set and is characterized by an efficient exploration strategy seeking for the optimal solution. The advantage of optimistic optimization is that one can guarantee bounds on the suboptimality with respect to the global optimum for a given computational budget. The 1-norm and ∞-norm objective functions often considered in model predictive control for continuous PWA systems are continuous PWA functions. We derive expressions for the core parameters required by optimistic optimization for the resulting optimization problem. By applying optimistic optimization, a sequence of control inputs is designed satisfying linear constraints. A bound on the suboptimality of the returned solution is also discussed. The performance of the proposed approach is illustrated with a case study on adaptive cruise control.

Receding-horizon control for max-plus linear systems with discrete actions using optimistic planning

Authors: Lucian Bușoniu

This paper addresses the infinite-horizon optimal control problem for max-plus linear systems where the considered objective function is a sum of discounted stage costs over an infinite horizon. The minimization problem of the cost function is equivalently transformed into a maximization problem of a reward function. The resulting optimal control problem is solved based on an optimistic planning algorithm. The control variables are the increments of system inputs and the action space is discretized as a finite set. Given a finite computational budget, a control sequence is returned by the optimistic planning algorithm. The first control action or a subsequence of the returned control sequence is applied to the system and then a receding-horizon scheme is adopted. The proposed optimistic planning approach allows us to limit the computational budget and also yields a characterization of the level of near-optimality of the resulting solution. The effectiveness of the approach is illustrated with a numerical example. The results show that the optimistic planning approach results in a lower tracking error compared with a finite-horizon approach when a subsequence of the returned control sequence is applied.

Robust Pallet Detection for Automated Logistics Operations

Authors: Sergiu Nedevschi

A pallet detection system is presented which is designed for automated forklifts for logistics operations. The system performs stereo reconstruction and pallets are detected using a sliding window approach. In this paper we propose a candidate generation method and we introduce feature descriptors for grayscale images that are tailored to the current task. The features are designed to be invariant to certain types of illumination changes and are called normalized pair differences because of the formula involved in their calculation. Experimental results validate our approach on extensive real world data.

The Role of the Complex Extended Textural Microstructure Co-occurrence Matrix in the Unsupervised Detection of the HCC Evolution Phases, based on Ultrasound Images

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea

The hepatocellular carcinoma (HCC) is a frequent malignant liver tumour and one of the main causes of death. Detecting the HCC evolution phases is an important issue, aiming the early diagnosis of this tumour and patient monitoring with maximum accuracy. Our objective is to discover the evolution stages of HCC, through unsupervised classification techniques, using advanced texture analysis methods. In this work, we assessed the role that the Haralick features derived from the Complex Extended Textural Microstructure Co-occurrence Matrices (CETMCM) have in the unsupervised detection of the HCC evolution stages. A textural model for these phases was also generated. The obtained results were validated by supervised classifiers, well known for their performance, such as the Multilayer Perceptron (MLP), Support Vector Machines (SVM), respectively decision trees and they were also compared with the previously obtained results in this domain. The final classification accuracy was about 90%.

The role of the complex textural microstructure co-occurrence matrices, based on Laws’ features, in the characterization and recognition of some pathological structures, from ultrasound images

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea

The non-invasive diagnosis, based on ultrasound images, is a challenge in nowadays research. We develop computerized, texture-based methods, for automatic and computer assisted diagnosis, using the information obtained from ultrasound images. In this work, we defined the co-occurrence matrix of complex textural microstructures determined by using the Laws’ convolution filters and we experimented it in order to perform the characterization and recognition of some important anatomical and pathological structures, within ultrasound images. These structures were the colorectal tumors and the gingival sulcus, the properties of the latter being important concerning the diagnosis and monitoring of the periodontal disease. We determined the textural model of these structures, using the classical and the newly defined textural features. For the automatic recognition, we used powerful classifiers, such as the Multilayer Perceptron, the Support-Vector Machines, decision-trees based classifiers such as Random Forest and C4.5, respectively AdaBoost in combination with the C4.5 algorithm.

Superpixels in Pedestrian Detection from Stereo Images in Urban Traffic Scenarios

Authors: Sergiu Nedevschi

Advanced Texture Analysis Techniques for Building Textural Models, with Applications in the Study of the Pathology Evolution Stages, based on Ultrasound Images

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea

ANALYSIS AND IMPROVEMENT TECHNIQUES FOR THE TRANSFER FUNCTION OF A PLANAR LOW - PASS FILTER

Authors: Călin Munteanu, Vasile Topa

In this paper new techniques for the analysis and improvement of the transfer function of a low-pass filter made in planar electromagnetic technology are proposed.The low pass filter is a fundamental element of the filter used for conduction electromagnetic interferences suppression.For this purpose, in order to determine the filter's transfer function, 3D electromagnetic CAE (Computer Aided Engineering) are coupled with simulation software programs which use equivalent electric circuits in order to directly obtain the frequency response.In the filter construction process the performance improvement techniques, proposed by the authors, are taken into account.These improvement techniques address two major directions of action, namely parasitic effects minimization for the component elements and attenuation increase at high frequency by analyzing the HF losses for the entire filter structure.Following the obtained results it can be concluded that the efficiency was increased for the Optimal planar low-pass filter with 1.2 dB in comparison to the Filter based on the initial structure.

XFEM BASED ALGORITHM FOR NUMERICAL OPTIMIZATION OF CURRENT DENSITY IN ELECTROCHEMICAL APPLICATIONS

Authors: Vasile Topa, Călin Munteanu

This paper presents an advanced numerical algorithm for the optimization of the current density distribution in electrochemical applications.The optimization principle consists in the distortion of the electric field by an insulating shield properly positioned between the electrodes.Such a distortion influences the distribution of current density at the electrodes under consideration.The electrochemical problem is approached by a Laplace equation which is numerically solved using a mesh-free algorithm based on the eXtended Finite Element Method (XFEM) while the optimal position of the insulating shield is managed by a genetic algorithm (GA).In order to reduce the computational time, two integration techniques were implemented in the XFEM algorithm: Gauss quadrature for the elements intersected by the discontinuity and natural coordinates for the other elements.The computational environment based on XFEM and GA was implemented in MatLab software.It can solve two dimensional (2D) and three dimensional (3D) applications and it was demonstrated on a practical electrochemical application for the optimization of current density distribution at the neighborhood between a cathode electrode in line with an insulator.The standard current density deviation at the cathode electrode is considerably reduced for optimal shield position comparing to the initial shield position.The use of the combined integration technique in the XFEM framework leads to an overall 70% improvement of the computational time compared with the standard XFEM integration technique.

Enacting textual entailment and ontologies for automated essay grading in chemical domain

Authors: Adrian Petru Groza

We propose a system for automated essay grading using ontologies and textual entailment. The process of textual entailment is guided by hypotheses, which are extracted from a domain ontology. Textual entailment checks if the truth of the hypothesis follows from a given text. We enact textual entailment to compare students answer to a model answer obtained from ontology. We validated the solution against various essays written by students in the chemistry domain.

Information retrieval in folktales using natural language processing

Authors: Adrian Petru Groza

Our aim is to extract information about literary characters in unstructured texts. We employ natural language processing and reasoning on domain ontologies. The first task is to identify the main characters and the parts of the story where these characters are described or act. We illustrate the system in a scenario in the folktale domain. The system relies on a folktale ontology that we have developed based on Propp's model for folktales morphology.

Data structuring for the ontological modelling of wind energy systems

Authors: Adrian Petru Groza

Small wind projects encounter difficulties to be efficiently deployed, partly because wrong way data and information are managed. Ontologies can overcome the drawbacks of partially available, noisy, inconsistent, and heterogeneous data sources, by providing a semantic middleware between low level data and more general knowledge. In this paper, we engineer an ontology for the wind energy domain using description logic as technical instrumentation. We aim to integrate corpus of heterogeneous knowledge, both digital and human, in order to help the interested user to speed-up the initialization of a small-scale wind project. We exemplify one use case scenario of our ontology, that consists of automatically checking whether a planned wind project is compliant or not with the active regulations.

An Open-Source Graphical Composite Modeling Editor and Simulation Tool Based on FMI and TLM Co-Simulation

Authors: Adrian Pop

A common situation in industry is that a system model (here a composite model) is composed of several submodels which may have been developed using different tools.FMI is one important technology for exporting/importing models between tools and/or connecting them via co-simulation.TLM based modeling and co-simulation is another important technique for modeling, connecting, and simulation of especially mechanical systems, which is simple, numerically stable, and efficient.A number of toolspecific simulation models, such as Modelica models, SimuLink models, Adams models, BEAST models, etc., have successfully been connected and simulated using TLM based co-simulation.However, previously there was no general open source tool for creation, graphic editing, and simulation of composite models connected via FMI or TLM based co-simulation.In this paper we present a graphical composite model editor based on OpenModelica which is integrated with the OpenModelica and the SKF TLM co-simulation frameworks to support both FMI and TLM based composite model editing and simulation.The editor supports creating, viewing and editing a composite model both in textual and graphical representation.The system supports simulation of composite models consisting of sub-models created using different tools.

Automatic Regression Testing of Simulation Models and Concept for Simulation of Connected FMUs in PySimulator

Authors: Adrian Pop

The Modelica and FMI tool ecosystem is growing each year with new tools and methods becoming available. The open Modelica standard promises portability but it is important to ensure that a certain model behaves the same in different Modelica tools or in a different version of the same tool. It is also very important (for model evolution) to check that a new version of the same model produces comparable results. Finally, it is desirable to verify that a model exported in FMU form from a Modelica tool gives exactly the same results as the original model. This paper presents a framework for automatic regression testing as part of PySimulator which provides an efficient and concise way of testing if a model or a range of models behaves in the same way in several tools or versions of a tool by checking that the results produced are essentially identical. The FMI standard has been adopted by many tool vendors and is growing in popularity each year. This paper proposes a concept for building and simulating a system made from connected FMUs generated by different tools. The FMUs for Co-Simulation can be connected together using a GUI. The system model built graphically in this way can be saved for later use or simulated directly inside PySimulator. Active development is going on to support simulation of connected FMUs for Model Exchange.

Flattening of Modelica State Machines: A Practical Symbolic Representation

Authors: Adrian Pop

Modelica 3.3 introduced dedicated built-in language support for state machines that was inspired by semantics known from Statechart and mode automata formalisms.The specification describes the semantics of these constructs in terms of data-flow equations that allows it to be related to the Modelica DAE representation which is the conceptual intermediate format of Modelica code after instance creation (flattening).However, a complete transformation of state machine constructs into data-flow equations at the stage of flattening requires an early commitment to implementation details that potentially hinders model optimizations at subsequent translation phases.Also, due to the required substantial model transformation the semantic distance between the original source model and the flattened representation is rather large.Hence, this paper proposes a more versatile symbolic representation for flattened state machine constructs that preserves the state machine's composition structure and allows postponing optimizations to subsequent compiler phases.

ALISA: An automatic lightly supervised speech segmentation and alignment tool

Authors: Adriana Stan

GEYSER: Enabling Green Data Centres in Smart Cities

Authors: Tudor Cioara, Anghel Ionuț

Vision and Control for UAVs: A Survey of General Methods and of Inexpensive Platforms for Infrastructure Inspection

Authors: Lucian Bușoniu

Unmanned aerial vehicles (UAVs) have gained significant attention in recent years. Low-cost platforms using inexpensive sensor payloads have been shown to provide satisfactory flight and navigation capabilities. In this report, we survey vision and control methods that can be applied to low-cost UAVs, and we list some popular inexpensive platforms and application fields where they are useful. We also highlight the sensor suites used where this information is available. We overview, among others, feature detection and tracking, optical flow and visual servoing, low-level stabilization and high-level planning methods. We then list popular low-cost UAVs, selecting mainly quadrotors. We discuss applications, restricting our focus to the field of infrastructure inspection. Finally, as an example, we formulate two use-cases for railway inspection, a less explored application field, and illustrate the usage of the vision and control techniques reviewed by selecting appropriate ones to tackle these use-cases. To select vision methods, we run a thorough set of experimental evaluations.

Near-Optimal Strategies for Nonlinear and Uncertain Networked Control Systems

Authors: Lucian Bușoniu

We consider problems where a controller communicates with a general nonlinear plant via a network, and must optimize a performance index. The system is modeled in discrete time and may be affected by a class of stochastic uncertainties that can take finitely many values. Admissible inputs are constrained to belong to a finite set. Exploiting some optimistic planning algorithms from the artificial intelligence field, we propose two control strategies that take into account the communication constraints induced by the use of the network. Both strategies send in a single packet long-horizon solutions, such as sequences of inputs. Our analysis characterizes the relationship between computation, near-optimality, and transmission intervals. In particular, the first strategy imposes at each transmission a desired near-optimality, which we show is related to an imposed transmission period; for this setting, we analyze the required computation. The second strategy has a fixed computation budget, and within this constraint it adapts the next transmission instant to the last state measurement, leading to a self-triggered policy. For this case, we guarantee long transmission intervals. Examples and simulation experiments are provided throughout the paper.

Machine Learning with Applications to Autonomous Systems

Authors: Lucian Bușoniu

Many acoustic channels suffer from interference which is neither narrowband nor impulsive. This relatively long duration partial band interference can be particularly detrimental to system performance. In operational networks, many “dropped” messages are lost due to partial band interference which corrupts different portions of the received signal depending on the relative position of the interferers, information source and receivers due to the slow speed of propagation. We survey recent work in interference mitigation as background motivation to develop a spatial diversity receiver for use in underwater networks and compare this multi-receiver interference mitigation strategy with a recently developed single receiver interference mitigation algorithm using experimental data collected from the underwater acoustic network at the Atlantic Underwater Test and Evaluation Center. The results indicate that both mitigation strategies are effective: parameterized interference cancellation is most effective at moderate SIRs whereas spatial diversity reconstruction is effective and realizes the most gain at high SIRs. We also apply the parametized interference cancellation to the problem of estimating mutually interfering waveforms when it is desired to know both time domain signals and find that it effectively extracts both mutually interfering linear frequency modulated (LFM) and orthogonal frequency division multiplexing (OFDM) waveforms.

Topology-preserving flocking of nonlinear agents using optimistic planning

Authors: Lucian Bușoniu

Classification of the liver tumors Using Co-Occurrence Matrices of textural Microstructures

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea

Experimental Reliability Improvement of Power Devices Operated Under Fast Thermal Cycling

Authors: Vasile Topa

In automotive applications, the lifetime of the power transistors is limited by the number of power pulses, induced by, e.g., short-circuit or inductive clamping events. This letter presents a solution for reliability improvement of double-diffused metal-oxide-semiconductor transistors which operate under thermomechanical stresses generated during power cycling.

Simulation and Control of the Vehicles Movement in the Case of the Overtaking Procedures

Authors: Adrian Petru Groza

In this paper, a solution for the simulation and control of the vehicles movement in the case of performing the overtaking procedures is proposed. In order to design the simulation system, the mathematical modeling of the vehicle movement is made, the obtained model being based on the state space representation. Also, the real-time working of the simulator is based both on the information obtained through the usage of a communication network between the vehicles and of the GPS-measured position. The main advantage of the proposed system consists in assisting the driver to perform a correct and a safe overtaking. Another original element presented in this paper is the including of the vehicle movement model in a control structure, resulting the possibility that the overtaking to be performed with a higher efficiency. The mathematical model, respectively the simulation system were tested and validated through some simulations associated to an overtaking scenario, simulations presented in this paper, too.

Pushing the Boundaries of Tractable Ontology Reasoning

Authors: Cristina Feier

LELA - A natural language processing system for Romanian tourism

Authors: Adrian Petru Groza, Anca Nicoleta Mărginean, Radu-Răzvan Slăvescu

Abstract—This paper presents a commercial semantic-based system for the Romanian tourism. The Lela system exploits both open linked data from Romanian and international sources, and also proprietary databases in the tourism domain. We present the process of creating the linked data set, based on: i) engineering the LELA Romanian tourism ontology, and ii) populating the ontology by linking open data. The system also provides a natural language interface for the Romanian language. The queries are automatically translated into SPARQL based on a controlled vocabulary derived from the Lela ontology. Index Terms—Semantic information retrieval, Query inter-faces, Natural language processing, Linked Data, Tourism on-tology

A formal approach for identifying assurance deficits in unmanned aerial vehicle software

Authors: Adrian Petru Groza

.While formal methods have proved to be unfeasible for large scale systems, argument-based safety cases offer a plausible alternative basis for certification of critical software. Our proposed method for increasing safety combines formal methods with argumentation-based reasoning. In a first step, we provide a formal representation of the the argumentative-based Goal Structuring Notation (GSN) standard used in industry. In a second step, our solution exploits reasoning in description logic to identify assurance deficits in the GSN model. The identified flaws are given to a hybrid logic-based model checker to be validated against a Kripke model. The method is illustrated for an unmanned aerial vehicle software, with reasoning performed in RacerPro engine and the HLMC model checker based on hybrid logic.

An Ontology Selection and Ranking System Based on the Analytic Hierarchy Process

Authors: Adrian Petru Groza

Selecting the desired ontology from a collection of available ones is essential for ontology reuse. We address the problem of evaluating, ranking and selecting ontologies according to user preferences. We exploit the Analytic Hierarchy Process (AHP) to solve the multiple-criteria decision problem and to model the preferences of the users. We use AHP to analyze the available ontologies from different perspectives and at different abstraction levels. The decision is based on the concrete end-node measurements and their relative importance at higher levels. For supporting the selection decision, we developed an ontology representation, reasoning and management system. The system applies different metrics on ontologies in order to feed the Analytic Hierarchy Process with facts. The running scenario applies our method to the task of reusing ontologies from the tourism domain.

A second order-cone programming relaxation for facility location problem

Authors: Adrian Petru Groza

In this paper we discuss the numerical solving of one of the well known problems related to the production systems. Namely, it is about facility location problem, which is reformulated in terms of non-convex minimization problems with quadratic constraints. After relaxation of the above problem, a model based on second-order cone programming is obtained.

Assuring safety in an air traffic control system with defeasible logic programming

Authors: Adrian Petru Groza

Assuring safety in complex technical systems is a crucial issue in several critical applications like air traffic control or medical devices. We present a preliminary framework based on argumentation for assisting flight controllers to reach a decision related to safety constraints in an ever changing environment in which sensor data is gathered at real time.

Numerical Simulation and Automatic Control of the pH Value in an Industrial Blunting System

Authors: Adrian Petru Groza

A solution for the pH control of the residual water in an industrial blunting system is proposed. The technological process associated to the blunting system is decomposed in four sub-processes connected in series and in parallel, each of them being a distributed parameter one. The mathematical models of the sub-processes are expressed using partial differential equations. Both this procedure and the advanced structure of the control system generate very high control performances. For the numerical simulation of the control system, a numerical method based on the Matrix of Partial Derivatives of the State Vector, associated with Taylor series is proposed. This method permits the numerical simulation of the systems that include in their structure distributed parameter processes. The conducted simulations proved high accuracy of our original method.

An Approach to Schedule Production using the Reservation Tables

Authors: Adrian Petru Groza

This paper deals with the problems of scheduling of manufacturing where specific requirements have to be considered like production of microwire and specialized products based on it. Scheduling of such production systems is very nontrivial because of some technological, human, and organizational issues. To tackle the respective issues some predictive and reactive scheduling schemes must be developed. The paper discusses the possibility to develop schedules based on the method of reservation tables, which is a well known method in the computer pipelines. The proposed approach it seems to be an acceptable one. In our opinion, the method of reservation tables may be successfully adapted and developed to schedule nontrivial production systems, especially when multi-mixed model lines work in small batches. Combining with other methods, the proposed approach may offer feasible schedules in computation times short enough to be accepted in real-life decision support systems.

ARGSAFE: usando argumentación para garantizar seguridaden Sistemas Técnicos Complejos

Authors: Adrian Petru Groza

En este trabajo, se presenta la problemática asociada al Proyecto Bilateral MINCYT-MECTS Argentina- Rumania llamado ARGSAFE que tiene como objetivo estudiar cómo utilizar argumentación rebatible para garantizar seguridad en sistemas técnicos complejos. Los resultados de la investigación se espera que sean aplicables a dominios tales como manejo de autos autónomos, aviación, ciencias médicas y redes de computadores.

Integrated Debugging of Modelica Models

Authors: Adrian Pop

The high abstraction level of equation-based object-oriented (EOO) languages such as Modelica has the drawback that programming and modeling errors are often hard to find.In this paper we present integrated static and dynamic debugging methods for Modelica models and a debugger prototype that addresses several of those problems.The goal is an integrated debugging framework that combines classical debugging techniques with special techniques for equation-based languages partly based on graph visualization and interaction.To our knowledge, this is the first Modelica debugger that supports both equation-based transformational and algorithmic code debugging in an integrated fashion.

Verification and Design Exploration through Meta Tool Integration with OpenModelica

Authors: Adrian Pop

Modelica models are typically used for simulation to investigate properties of a possible system designs.This is often done manually or combined with optimization to select the best design parameters.It is desirable to have systematic and partly automated support for exploration of the design space of possible designs and verifying their properties vs. requirements.The META design tool chain is being developed to support this goal.It provides an integration framework for components, designs, design spaces, requirements, and test benches, as well as verification of requirements for the generated design models during design explorationThis paper gives an overview of the META tools and their integration with OpenModelica.The integrated environment currently has four main uses of OpenModelica: importing Modelica models into the META tool model structure, performing simulations within test benches, analyzing Modelica models and automatically adding fault modes, and extracting equations (DAEs) for formal verification tools, e.g. the QRM using qualitative reasoning.A prototype of the integrated tool framework is in operation, being able to generate and simulate thousands of designs in an automated manner.

Bootstrapping a Compiler for an Equation-Based Object-Oriented Language

Authors: Adrian Pop

What does it mean to bootstrap a compiler, and why do it?This paper reports on the first bootstrapping of a full-scale EOO (Equation-based Object-Oriented) modeling language such as Modelica.Bootstrapping means that the compiler of a language can compile itself.However, the usual application area for the Modelica is modeling and simulation of complex physical systems.Fortunately it turns out that with some minor extensions, the Modelica language is well suited for the modeling of language semantics.We use the name MetaModelica for this slightly extended Modelica.This is a prerequisite for bootstrapping which requires that the language can be used to model and/or implement itself.The OpenModelica Compiler (OMC) has been written in this MetaModelica language.It originally supported only the standard Modelica language but has been gradually extended to also cover the MetaModelica language extensions.After substantial work, OMC is able to quickly compile itself and produces an executable with good performance.The benefits include a more extensible and maintainable compiler by introducing improved language constructs and a more powerful runtime that makes it easy to add functionality such as parser generators, debuggers, and profiling tools.Future work includes extracting and restructuring parts of OMC, making the compiler smaller and more modular and extensible.This will also make it easier to interface with OMC, making it possible to create more powerful and user-friendly OpenModelica-based tools.The compiler and its bootstrapping is a major effort -it is currently about 330 000 lines of code, and the MetaModelica extensions are used routinely by approximately ten developers on a daily basis.

Integrated Debugging of Equation-Based Models

Authors: Adrian Pop

The high abstraction level of equation-based objectoriented languages (EOO) such as Modelica has the drawback that programming and modeling errors are often hard to find.In this paper we present the first integrated debugger for equation-based languages like Modelica, which can combine static and dynamic methods for run-time debugging of equation-based Modelica models during simulations.This builds on and extends previous results from a transformational static equation debugger and a dynamic debugger for the algorithmic subset of Modelica.

Neural net word representations for phrase-break prediction without a part of speech tagger

Authors: Adriana Stan, Mircea Giurgiu

The use of shared projection neural nets of the sort used in language modelling is proposed as a way of sharing parameters between multiple text-to-speech system components. We experiment with pretraining the weights of such a shared projection on an auxiliary language modelling task and then apply the resulting word representations to the task of phrase-break prediction. Doing so allows us to build phrase-break predictors that rival conventional systems without any reliance on conventional knowledge-based resources such as part of speech taggers.

$\mathcal{EL}$ -ifying Ontologies

Authors: Cristina Feier

Is Your Ontology as Hard as You Think? Rewriting Ontologies into Simpler DLs

Authors: Cristina Feier

Abstract. We investigate cases where an ontology expressed in a seem-ingly hard DL can be polynomially reduced to one in a simpler logic, while preserving reasoning outcomes for classification and fact entail-ment. Our transformations target the elimination of inverse roles, uni-versal and existential restrictions, and in the best case allow us to rewrite the given ontology into one of the OWL 2 profiles. Even if an ontology cannot be fully rewritten into a profile, in many cases our transforma-tions allow us to exploit further optimisation techniques. Moreover, the elimination of some out-of-profile axioms can improve the performance of modular reasoners, such as MORe. We have tested our techniques on both classification and data reasoning tasks with encouraging results.

Reasoning with forest logic programs

Authors: Cristina Feier

Answer Set Programming (ASP) ist ein beliebter, regelbasierter Formalismus zur Beschreibung und Lösung von kombinatorischen Suchproblemen. Open Answer Set Programming (OASP) erweitert Answer Set Programming um eine Semantik mit offenen Domänen: Programme erhalten eine Interpretation in beliebigen Dömanen, die anonyme Individuen enthalten kann, d.h. Individuen, die nicht explizit im Programm aufscheinen. Dies ermöglicht es mittels OASP generisches Wissen zu formulieren. Gleichzeitig erbt OASP von ASP die nicht-monotone Behandlung der Negation unter der stabilen Modell-Semantik (Stable Model Semantics). OASP verbindet auf seine Art zwei wichtige Paradigmen der Wissensrepräsentation: die klassische Prädikatenlogik erster Stufe (First-Order Logic, FOL) und die Welt der nicht-monotonen Regeln. Im allgemeinen Fall ist OASP unentscheidbar. Um Entscheidbarkeit zu erreichen wurden deshalb verschiedene Fragmente von OASP durch deshalb syntaktische Einschränkungen der Form der Regeln definiert. Beispiele solcher Fragmente sind Conceptual Logic Programs (CoLPs) - ein Fragment, das die Verwendung von Konstanten verbietet und die Tree-Modell Eigenschaft erfüllt -, und Forest Logic Programs (FoLPs) - ein Fragment mit der Forest-Modell Eigenschaft. CoLPs können Schließen (engl. Reasoning) in der Beschreibungslogik (engl. Description Logic) SHQ simulieren und es wurde durch eine Reduktion auf two-way alternating tree automata gezeigt, dass sie entscheidbar sind. FoLPs hingegen können Reasoning in der Beschreibungslogik SHOQ simulieren, aber es war vor dieser Arbeit nicht bekannt, ob sie entscheidbar sind oder nicht. Entscheidbarkeit war vorher nur für Fragmente von azyklischen und lokalen FOLPs bewiesen, in denen die Verwendung von Rekursion sehr stark eingeschränkt ist. Deswegen ist der Reasoning-Support für CoLPs und FoLPs ziemlich begrenzt. Die Forest-Modell Eigenschaft von FOLPs eröffnet die Möglichkeit, sie mit tableau-basierten Methoden zu untersuchen, die Modelle mit Forest Form erzeugen. Im Gegensatz zur Reduktion auf Automaten, sind zielorientierte Prozeduren wie Tableau-Algorithmen nicht worst-case optimal, aber sie eignen sich für Optimierungen und verhalten sich in der Praxis gut. Daher ist das Ziel der Arbeit, tableau-basierte Algorithmen für das Reasoning mit FoLPs zu entwickeln. Die Wahl von FoLPs erfolgte, weil sie ein sehr ausdrucksstarkes Fragment von OASP sind, das auch CoLPs umfasst. Eine der größten Herausforderungen bei der Entwicklung von Tableau-Algorithmen für FoLPs besteht darin, für jedes Atom im konstruierten Modell sicher zu stellen, dass es "well-supported" ist, d.h. dass es eine Begründung für das Atom im Modell gibt, die weder zirkulär noch unendlich ist. Die Erweiterung des Standard Blocking, wie es von Tableau Algorithmen angewandt wird, um die Überprüfung von Well-Supportedness ist aber nicht ausreichend, um die Termination des Verfahrens zu garantieren. Deswegen stützen sich unsere Algorithmen stark auf Techniken mit denen für FoLPs die Finite-Modell Eigenschaft erreichtwerden kann; diese werden dann in Laufzeit-Endbedingungen umgewandelt. Basierend auf zwei solche Techniken entwickeln wir Algorithmen für die Überprüfung der Erfüllbarkeit (engl. Satisfiability) für unäre Prädikate bezüglich FoLPs, die im schlechtesten Fall nichtdeterministisch doppelt exponentielle bzw. nichtdeterministisch einfach exponentielle Laufzeit haben. Während die zweite Klasse von Algorithmen ein besseres Laufzeitverhalten im schlechtesten Fall aufweist, erfordert sie ausführliche Aufzeichnungen (engl. Bookkeeping). Wir identifizieren ein neues Fragment von FoLPs, simple Forest Logic Programs, das azyklische FoLPs verallgemeinert und für die man einen unkomplizierten nicht-deterministischen Algorithmus entwickeln kann. Die oben erwähnten Algorithmen zeigen, dass Satisfiability-Testen von unären Prädikaten in FoLPs in nichtdeterministische einfach exponentieller Zeit entschieden werden kann. Aufgrund der Tatsache, dass FoLPs Reasoning in der Description Logic SHOQ simulieren können, wissen wir, dass Reasoning in FolPs EXPTIME hart ist. Auf der Suche nach worst-case optimalen Algorithmen untersuchen wir die Determinisierung der oben erwähnten Algorithmen mit Hilfe von AND/OR Datenstrukturen. Wir erhalten zwei worst-case optimale Algorithmen für das Reasoning mit Fragmenten von CoLPs und einfachen FoLPs.

Consensus for black-box nonlinear agents using optimistic optimization

Authors: Lucian Bușoniu

Stability of infinite-horizon optimal control with discounted cost

Authors: Lucian Bușoniu

We investigate the stability of general nonlinear discrete-time systems controlled by an optimal sequence of inputs that minimizes an infinite-horizon discounted cost. We first provide conditions under which a global asymptotic stability property is ensured for the corresponding undiscounted problem. We then show that this property is semiglobally and practically preserved in the discounted case, where the adjustable parameter is the discount factor. We then focus on a scenario where the stage cost is bounded and we explain how our framework applies to guarantee stability in this case. Finally, we provide sufficient conditions, including boundedness of the stage cost, under which the value function, which serves as a Lyapunov function for the analysis, is continuous. As already shown in the literature, the continuity of the Lyapunov function is crucial to ensure some nominal robustness for the closed-loop system.

Semi-Supervised Segmentation of Ultrasound Images Based on Patch Representation and Continuous Min Cut

Authors: Sergiu Nedevschi

Ultrasound segmentation is a challenging problem due to the inherent speckle and some artifacts like shadows, attenuation and signal dropout. Existing methods need to include strong priors like shape priors or analytical intensity models to succeed in the segmentation. However, such priors tend to limit these methods to a specific target or imaging settings, and they are not always applicable to pathological cases. This work introduces a semi-supervised segmentation framework for ultrasound imaging that alleviates the limitation of fully automatic segmentation, that is, it is applicable to any kind of target and imaging settings. Our methodology uses a graph of image patches to represent the ultrasound image and user-assisted initialization with labels, which acts as soft priors. The segmentation problem is formulated as a continuous minimum cut problem and solved with an efficient optimization algorithm. We validate our segmentation framework on clinical ultrasound imaging (prostate, fetus, and tumors of the liver and eye). We obtain high similarity agreement with the ground truth provided by medical expert delineations in all applications (94% DICE values in average) and the proposed algorithm performs favorably with the literature.

Assisting Navigation in Homogenous Fog

Authors: Sergiu Nedevschi

An important cause of road accidents is the reduced visibility due to the presence of fog or haze. For this reason, there is a fundamental need for Advanced Driving Assistance Systems (ADAS) based on efficient real time algorithms able to detect the presence of fog, estimate the fog's density, determine the visibility distance and inform the driver about the maximum speed that the vehicle should be traveling. Our solution is an improvement over existing methods of detecting fog due to the temporal integration of the horizon line and inflection point in the image. Our method performs in real time; approximately 50 frames per second. It is based on a single in-vehicle camera and is able to detect day time fog in real time in a wide range of scenarios, including urban scenarios.

Multi-feature Real Time Pedestrian Detection from Dense Stereo SORT-SGM Reconstructed Urban Traffic Scenarios

Authors: Sergiu Nedevschi

In this paper, a real-time system for pedestrian detection in traffic scenes is proposed. It takes the advantage of having a pair of stereo video-cameras for acquiring the image frames and uses a sub-pixel level optimized semi-global matching (SORT-SGM) based stereo reconstruction for computing the dense 3D points map with high accuracy. A multiple paradigm detection module considering 2D, 3D and optical flow information is used for segmenting the candidate obstacles from the scene background. Novel features like texture dissimilarity, humans' body specific features, distance related measures and speed are introduced and combined in a feature vector with traditional features like HoG score, template matching contour score and dimensions. A random forest (RF) classifier is trained and then applied in each frame for distinguishing the pedestrians from other obstacles based on the feature vector. A k-NN algorithm on the classification results over the last frames is applied for improving the accuracy and stability of the tracked obstacles. Finally, two comparisons are made: first between the classification results obtained by using the new SORT-SGM and the older local matching approach for stereo reconstruction and the second by comparing the different features RF classification results with other classifiers' results.

Improving Filter Performances for Conducted Electromagnetic Interferences Suppression

Authors: Călin Munteanu, Vasile Topa

Improving the performances of a filter for conductive electromagnetic interferences suppression involves two major directions of action: the first direction refers to the increase of attenuation at high frequency, while the second direction refers to the suppression of the parasitic effects in the constitutive devices. Thus, in this light, the paper presents the authors’ contribution to the two major directions of actions mentioned above; in the first part, techniques for HF loss increase are presented, while in the second part techniques for minimizing the structural parasitic capacitance are shown. In the third part of the paper these techniques are applied simultaneously to an EMI filter made in planar electromagnetic technology in order to study its performance.

TUNDRA: a multilingual corpus of found data for TTS research created with light supervision

Authors: Adriana Stan

Simple4All Tundra (version 1.0) is the first release of a standardised multilingual corpus designed for text-to-speech re-search with imperfect or found data. The corpus consists of approximately 60 hours of speech data from audiobooks in 14 languages, as well as utterance-level alignments obtained with a lightly-supervised process. Future versions of the corpus will include finer-grained alignment and prosodic annotation, all of which will be made freely available. This paper gives a gen-eral outline of the data collected so far, as well as a detailed description of how this has been done, emphasizing the mini-mal language-specific knowledge and manual intervention used to compile the corpus. To demonstrate its potential use, text-to-speech systems have been built for all languages using unsu-pervised or lightly supervised methods, also briefly presented in the paper. Index Terms: multilingual corpus, light supervision, imperfect data, found data, text-to-speech, audiobook data

Enacting Social Argumentative Machines in Semantic Wikipedia

Authors: Adrian Petru Groza

This research advocates the idea of combining argumentation theory with the social web technology, aiming to enact large scale or mass argumentation. The proposed framework allows mass-collaborative editing of structured arguments in the style of semantic wikipedia. The long term goal is to apply the abstract machinery of argumentation theory to more practical applications based on human generated arguments, such as deliberative democracy, business negotiation, or self-care. The ARGNET system was developed based on ther Semantic MediaWiki framework and on the Argument Interchange Format (AIF) ontology.

OntoRich - A Support Tool for Semi-Automatic Ontology Enrichment and Evaluation

Authors: Adrian Petru Groza

This paper presents the OntoRich framework, a support tool for semi-automatic ontology enrichment and evaluation. The WordNet is used to extract candidates for dynamic ontology enrichment from RSS streams. With the integration of OpenNLP the system gains access to syntactic analysis of the RSS news. The enriched ontologies are evaluated against several qualitative metrics.

Designing Electronic Markets for Defeasible-based Contractual Agents

Authors: Adrian Petru Groza

The design of punishment policies applied to specific domains linking agents actions to material penalties is an open research issue. The proposed framework applies principles of contract law to set penalties: expectation damages, opportunity cost, reliance damages, and party design remedies. In order to decide which remedy provides maximum welfare within an electronic market, a simulation environment called DEMCA (Designing Electronic Markets for Contractual Agents) was developed. Knowledge representation and the reasoning capabilities of the agents are based on an extended version of temporal defeasible logic.

Justificatory and Explanatory Argumentation for Committing Agents

Authors: Adrian Petru Groza

In the interaction between agents we can have an explicative discourse, when communicating preferences or intentions, and a normative discourse, when considering normative knowledge. For justifying their actions our agents are endowed with a Justification and Explanation Logic (JEL), capable to cover both the justification for their commitments and explanations why they had to act in that way, due to the current situation in the environment. Social commitments are used to formalise justificatory and explanatory patterns. The combination of ex- planation, justification, and commitments

Automating the Dispute Resolution in Task Dependency Network

Authors: Adrian Petru Groza

When perturbation or unexpected events do occur, agents need protocols for repairing or reforming the supply chain. Unfortunate contingency could increase too much the cost of performance, while breaching the current contract may be more efficient. In our framework the principles of contract law are applied to set penalties: expectation damages, opportunity cost, reliance damages, and party design remedies, and they are introduced in the task dependency model

Unsupervised and lightly-supervised learning for rapid construction of TTS systems in multiple languages from 'found' data: evaluation and analysis

Authors: Adriana Stan, Mircea Giurgiu

This paper presents techniques for building text-to-speech front-ends in a way that avoids the need for language-specific expert knowledge, but instead relies on universal resources (such as the Unicode character database) and unsupervised learning from unannotated data to ease system development. The acquisition of expert language-specific knowledge and expert annotated data is a major bottleneck in the development of corpus-based TTS systems in new languages. The methods presented here side-step the need for such resources as pronunciation lexicons, phonetic feature sets, part of speech tagged data, etc. The paper explains how the techniques introduced are applied to the 14 languages of a corpus of `found' audiobook data. Results of an evaluation of the intelligibility of the systems resulting from applying these novel techniques are presented.

Lightly supervised GMM VAD to use audiobook for speech synthesiser

Authors: Adriana Stan

Audiobooks have been focused on as promising data for training Text-to-Speech (TTS) systems. However, they usually do not have a correspondence between audio and text data. Moreover, they are usually divided only into chapter units. In practice, we have to make a correspondence of audio and text data before we use them for building TTS synthesisers. However aligning audio and text data is time-consuming and involves manual labor. It also requires persons skilled in speech processing. Previously, we have proposed to use graphemes for automatically aligning speech and text data. This paper further integrates a lightly supervised voice activity detection (VAD) technique to detect sentence boundaries as a pre-processing step before the grapheme approach. This lightly supervised technique requires time stamps of speech and silence only for the first fifty sentences. Combining those, we can semi-automatically build TTS systems from audiobooks with minimum manual intervention. From subjective evaluations we analyse how the grapheme-based aligner and/or the proposed VAD technique impact the quality of HMM-based speech synthesisers trained on audiobooks.

The Simple4All entry to the Blizzard Challenge 2013

Authors: Adriana Stan

We describe the synthetic voices entered into the 2013 Blizzard Challenge by the SIMPLE 4 ALL consortium.The 2013 Blizzard Challenge presents an opportunity to test and benchmark some of the tools we have been developing to address two problems of interest: 1) how best to learn from plentiful 'found' data, and 2) how to produce systems in arbitrary new languages with minimal annotated data and language-specific expertise on the part of the system builders.We here explain how our tools were used to address these problems on the different tasks of the challenge, and provide some discussion of the evaluation results.

Evaluation of sentiment polarity prediction using a dimensional and a categorical approach

Authors: Adriana Stan, Mircea Giurgiu, Rodica Potolea

In this paper we evaluate two approaches for predicting the sentiment polarity of an utterance. The first method is based on a 3-dimensional model which takes into account text expressiveness in terms of valence, arousal and dominance. The second one determines the word's semantic orientation according to Chi-square and Relevance factor statistic metrics. We describe the general flow of the methods and their extracted features, as well as their predictability potential using different machine learning algorithms, Naïve Bayes, SVM and C4.5. The evaluation is performed on four emotional datasets: Semeval 2007 “Affective Text”, ISEAR (International Survey on Emotional Antecedents and Reactions), children's fairy-tales and a movie review dataset. The results show a high correlation of the prediction performance with the database content, as well as to the average number of words within the classified text instances.

A Swarm-inspired Technique for Self-Organizing and Consolidating Data Centre Servers

Authors: Tudor Cioara, Anghel Ionuț

This paper proposes a swarm-inspired data centre self-organizing and consolidation technique which aims at reducing the power demand in data centres while ensuring the workload execution within the established performance parameters.Each data centre server is managed by an intelligent agent that implements a bird's migration-inspired behaviour to decide on the appropriate server consolidation actions.The selected actions are executed to achieve an optimal utilization of server computing resources thus lowering power demand.The data centre servers self-organize in logical clusters according to the birds V-formation self-organizing migration model.The results are promising showing that by using the proposed swarm-inspired solution, the data centre Deployed Hardware Utilization Ratio Green Indicator increases compared to the widely used Fit First consolidation algorithm.The average power saving of the proposed technique is around 40% of the power demanded by the data centre computing resources and about 16% of its total power demand including the IT facility, when comparing to OpenNebula Fit First consolidation technique.This paper is an extended version of the one published in WIMS'12 proceedings showing more details about the swarm-inspired consolidation technique and the defined algorithms.

Consensus for agents with general dynamics using optimistic optimization

Authors: Lucian Bușoniu

An important challenge in multiagent systems is consensus, in which the agents must agree on certain controlled variables of interest. So far, most consensus algorithms for agents with nonlinear dynamics exploit the specific form of the nonlinearity. Here, we propose an approach that only requires a black-box simulation model of the dynamics, and is therefore applicable to a wide class of nonlinearities. This approach works for agents communicating on a fixed, connected network. It designs a reference behavior with a classical consensus protocol, and then finds control actions that drive the nonlinear agents towards the reference states, using a recent optimistic optimization algorithm. By exploiting the guarantees of optimistic optimization, we prove that the agents achieve practical consensus. A representative example is further analyzed, and simulation results on nonlinear robotic arms are provided.

A general approach for consensus using optimistic planning

Authors: Lucian Bușoniu

An important challenge in multiagent systems is consensus, in which the agents are required to synchronize certain controlled variables of interest, often using only an incomplete and time-varying communication graph. We propose a consensus approach based on optimistic planning (OP), a predictive control algorithm that finds near-optimal control actions for general dynamics and reward functions (costs). At every step, each agent uses OP to solve a local control problem with rewards that express the consensus objectives. Neighboring agents coordinate by exchanging their predicted behaviors in a predefined order. Due to its generality, OP consensus can adapt to any agent dynamics and, by changing the reward function, to a variety of consensus objectives. While theoretical analysis is still open, OP consensus is demonstrated in experiments for two problems. The first problem is velocity consensus (flocking) with a time-varying communication graph, where OP preserves connectivity better than a classical algorithm. The second problem is the leaderless and leader-based consensus of robotic arms, where OP easily deals with the nonlinear dynamics.

Planification Optimiste dans les Processus Décisionnels de Markov avec Croyance

Authors: Lucian Bușoniu

peer reviewed

Optimistic Planning for Belief-Augmented Markov Decision Processes

Authors: Lucian Bușoniu

Abstract—This paper presents the Bayesian Optimistic Plan-ning (BOP) algorithm, a novel model-based Bayesian reinforce-ment learning approach. BOP extends the planning approach of the Optimistic Planning for Markov Decision Processes (OP-MDP) algorithm [10], [9] to contexts where the transition model of the MDP is initially unknown and progressively learned through interactions within the environment. The knowledge about the unknown MDP is represented with a probability distribution over all possible transition models using Dirichlet dis-tributions, and the BOP algorithm plans in the belief-augmented state space constructed by concatenating the original state vector with the current posterior distribution over transition models. We show that BOP becomes Bayesian optimal when the budget parameter increases to infinity. Preliminary empirical validations show promising performance. I.

Low Complexity Approach for High Throughput Belief-Propagation based Decoding of LDPC Codes

Authors: Sergiu Nedevschi

The paper proposes a low complexity belief propagation (BP) based decoding algorithm for LDPC codes. In spite of the iterative nature of the decoding process, the proposed algorithm provides both reduced complexity and increased BER performances as compared with the classic min-sum (MS) algorithm, generally used for hardware implementations. Linear approximations of check-nodes update function are used in order to reduce the complexity of the BP algorithm. Considering this decoding approach, an FPGA based hardware architecture is proposed for implementing the decoding algorithm, aiming to increase the decoder throughput. FPGA technology was chosen for the LDPC decoder implementation, due to its parallel computation and reconfiguration capabilities. The obtained results show improvements regarding decoding throughput and BER performances compared with state-of-the-art approaches.

STUDIES OF INDUCTANCE VARIATION FOR SQUARE SPIRAL INDUCTORS USING CIBSOC SOFTWARE

Authors: Vasile Topa, Călin Munteanu

This paper presents the most recent developments regarding the square spiral inductors inductance evaluation.The study was performed in order to develop, implement and validate powerful software, CIBSOC able to determine the inductances of spiral inductors and also to optimize the spiral inductors layouts.We have studied the inductance variation of square spiral inductors.The four modules of the CIBSOC software were described, applied for the computation of inductance and the optimization of spiral inductors layout.Also, we have analyzed an extensive range of square spiral inductors using CIBSOC and compared the results of our analyses with the simulations performed using a commercial solver in order to validate our program.Our data demonstrated that the analyses carried out with CIBSOC have a high accuracy, so that the program is very useful for the spiral inductors design.Also the analyses are solved in short running times and using low computational resources.It is also useful for the spiral inductors optimization, because it affords optimal solutions for spiral inductor layouts in terms of the technological limits and/or of the users' needs.The program has a friendly interface, being easy to use its computation and optimization modules and the running times and computational resources are smaller when compared with analogous tools.

MITIGATION OF POWER FREQUENCY MAGNETIC FIELD NEARBY POWER LINES USING RECTANGULAR FRAMES

Authors: Călin Munteanu, Vasile Topa

The paper outlines a technique to mitigate the power frequency magnetic field produced in its vicinity by the transport and distribution power lines.The technique consists in introduction below the line conductors of a passive loop of rectangular shape.Due to the induced currents in the rectangular frame, the total magnetic field below this structure is significantly reduced.The method's efficiency depends by the line conductor's configuration and also by the dimensions of the frame.The efficiency of the method can be improved by adding external capacitances to the passive loop.These external capacitances can lead to series resonance effect in the loop equivalent circuit that increases the value of the induced current flowing in the frame but they also enhance the induced current phase.Thus, in the first part of the paper the theoretical background is emphasized.Starting from Maxwell laws the expressions of the induced electromotive force in the frame structure and the corresponding current flowing through it are determined.Then the magnetic field computation formulae are outlined.Using these formulae one can compute the magnetic field values in points located below the power line, emphasizing the effect of the presence or not of the protective frame.In the second part of the paper two conductor line arrangements are used in order to proof the efficiency of the mitigation technique proposed.The improvement in the field mitigation by the usage of an external capacitance is also made.

Steady State Analysis of Thin Oxide Barrier Layers Formed in Neutral and Weakly Alkaline Solutions

Authors: Vasile Topa

Abstract not Available.

A Survey of Actor-Critic Reinforcement Learning: Standard and Natural Policy Gradients

Authors: Lucian Bușoniu

Policy-gradient-based actor-critic algorithms are amongst the most popular algorithms in the reinforcement learning framework. Their advantage of being able to search for optimal policies using low-variance gradient estimates has made them useful in several real-life applications, such as robotics, power control, and finance. Although general surveys on reinforcement learning techniques already exist, no survey is specifically dedicated to actor-critic algorithms in particular. This paper, therefore, describes the state of the art of actor-critic algorithms, with a focus on methods that can work in an online setting and use function approximation in order to deal with continuous state and action spaces. After starting with a discussion on the concepts of reinforcement learning and the origins of actor-critic algorithms, this paper describes the workings of the natural gradient, which has made its way into many actor-critic algorithms over the past few years. A review of several standard and natural actor-critic algorithms is given, and the paper concludes with an overview of application areas and a discussion on open issues.

Plausible Description Logic Programs for Stream Reasoning

Authors: Adrian Petru Groza

Sensor networks are estimated to drive the formation of the future Internet, with stream reasoning responsible for analysing sensor data. Stream reasoning is defined as real time logical reasoning on large, noisy, heterogeneous data streams, aiming to support the decision process of large numbers of concurrent querying agents. In this research we exploited non-monotonic rule-based systems for handling inconsistent or incomplete information and also ontologies to deal with heterogeneity. Data is aggregated from distributed streams in real time and plausible rules fire when new data is available. The advantages of lazy evaluation on data streams were investigated in this study, with the help of a prototype developed in Haskell.

DESCRIPTION PLAUSIBLE LOGIC PROGRAMS FOR STREAM REASONING

Authors: Adrian Petru Groza

Static and Dynamic Debugging of Modelica Models

Authors: Adrian Pop

The high abstraction level of equation-based objectoriented languages (EOO) such as Modelica has the drawback that programming and modeling errors are often hard to find.In this paper we present static and dynamic debugging methods for Modelica models and a debugger prototype that addresses several of those problems.The goal is an integrated debugging framework that combines classical debugging techniques with special techniques for equation-based languages partly based on graph visualization and interaction.To our knowledge, this is the first Modelica debugger that supports both transformational and algorithmic code debugging.

Efficient Debugging of Large Algorithmic Modelica Applications

Authors: Adrian Pop

A grapheme-based method for automatic alignment of speech and text data

Authors: Adriana Stan

This paper introduces a method for automatic alignment of speech data with unsynchronised, imperfect transcripts, for a domain where no initial acoustic models are available. Using grapheme-based acoustic models, word skip networks and orthographic speech transcripts, we are able to harvest 55% of the speech with a 93% utterance-level accuracy and 99% word accuracy for the produced transcriptions. The work is based on the assumption that there is a high degree of correspondence between the speech and text, and that a full transcription of all of the speech is not required. The method is language independent and the only prior knowledge and resources required are the speech and text transcripts, and a few minor user interventions.

Optimistic Planning for Markov Decision Processes

Authors: Lucian Bușoniu

International audience

Imitation learning with non-parametric regression

Authors: Lucian Bușoniu

Humans are very fast learners. Yet, we rarely learn a task completely from scratch. Instead, we usually start with a rough approximation of the desired behavior and take the learning from there. In this paper, we use imitation to quickly generate a rough solution to a robotic task from demonstrations, supplied as a collection of state-space trajectories. Appropriate control actions needed to steer the system along the trajectories are then automatically learned in the form of a (nonlinear) state-feedback control law. The learning scheme has two components: a dynamic reference model and an adaptive inverse process model, both based on a data-driven, non-parametric method called local linear regression. The reference model infers the desired behavior from the demonstration trajectories, while the inverse process model provides the control actions to achieve this behavior and is improved online using learning. Experimental results with a pendulum swing-up problem and a robotic arm demonstrate the practical usefulness of this approach. The resulting learned dynamics are not limited to single trajectories, but capture instead the overall dynamics of the motion, making the proposed approach a promising step towards versatile learning machines such as future household robots, or robots for autonomous missions.

Abdominal Tumor Characterization and Recognition Using Superior-Order Cooccurrence Matrices, Based on Ultrasound Images

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea

The noninvasive diagnosis of the malignant tumors is an important issue in research nowadays. Our purpose is to elaborate computerized, texture-based methods for performing computer-aided characterization and automatic diagnosis of these tumors, using only the information from ultrasound images. In this paper, we considered some of the most frequent abdominal malignant tumors: the hepatocellular carcinoma and the colonic tumors. We compared these structures with the benign tumors and with other visually similar diseases. Besides the textural features that proved in our previous research to be useful in the characterization and recognition of the malignant tumors, we improved our method by using the grey level cooccurrence matrix and the edge orientation cooccurrence matrix of superior order. As resulted from our experiments, the new textural features increased the malignant tumor classification performance, also revealing visual and physical properties of these structures that emphasized the complex, chaotic structure of the corresponding tissue.

The Role of the Superior Order GLCM in the Characterization and Recognition of the Liver Tumors from Ultrasound Images

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea

The hepatocellular carcinoma (HCC) is the most frequent malignant liver tumor. It often has a similar visual aspect with the cirrhotic parenchyma on which it evolves and with the benign liver tumors. The golden standard for HCC diagnosis is the needle biopsy, but this is an invasive, dangerous method. We aim to develop computerized,noninvasive techniques for the automatic diagnosis of HCC, based on information obtained from ultrasound images. The texture is an important property of the internal organs tissue, able to provide subtle information about the pathology. We previously defined the textural model of HCC, consisting in the exhaustive set of the relevant textural features, appropriate for HCC characterization and in the specific values of these features. In this work, we analyze the role that the superior order Grey Level Cooccurrence Matrices (GLCM) and the associated parameters have in the improvement of HCC characterization and automatic diagnosis. We also determine the best spatial relations between the pixels that lead to the highest performances, for the third, fifth and seventh order GLCM. The following classes will be considered: HCC, cirrhotic liver parenchyma on which it evolves and benign liver tumors.

Iterative Methods for Obtaining Energy-Minimizing Parametric Snakes with Applications to Medical Imaging

Authors: Sergiu Nedevschi, Delia Alexandrina Mitrea

After a brief survey on the parametric deformable models, we develop an iterative method based on the finite difference schemes in order to obtain energy-minimizing snakes. We estimate the approximation error, the residue, and the truncature error related to the corresponding algorithm, then we discuss its convergence, consistency, and stability. Some aspects regarding the prosthetic sugical methods that implement the above numerical methods are also pointed out.

AN ALGORITHM FOR SOLVING SOME NONLINEAR SYSTEMS WITH APPLICATIONS TO EXTREMUM PROBLEMS

Authors: Sergiu Nedevschi

We consider a class of nonlinear systems for which a positive solution exists and is unique. Such systems appear quite naturally in several applications concerning difference equations. Moreover, certain extremum problems can be reduced to solving these systems. In order to solve such problems we develop a quasi-Newton algorithm which is very efficient just because the existence and uniqueness of the solution are guaranteed. Several numerical examples illustrate the general results.

Improving Stereo Vision Odometry for Mobile Robots in Outdoor Dynamic Environments

Authors: Sergiu Nedevschi

Vision based Real-time Modeling of Dynamic Unstructured Environments in Driving Scenarios

Authors: Sergiu Nedevschi

Testing Determination and Validation of the Mathematical Equation an Antimony PH Sensor

Authors: Vasile Topa

This paper presents testing mode, determinating and validating the mathematical equation an antimony pH sensor.The introduction offers an incursion on the mine water and a definition of the pH concept, a brief history of it and continues with a short presentation of the elements which can make a possible variant of pH measurement, more precise the variant which use pH electrode/pH sensor.In the content of the paper is exposed a theoretical synthesis of the way of testing and validation of pH electrodes/pH sensors, which makes the transition to the actual testing of the studied pH sensor.The elements used for test and also the sensor response are viewed in photographs and graphics, which helps the reader to understand the practical part.Further, in paper, are presented short theoretical considerations followed by the dependence of electrical potential sensor analysis in relation with the change in pH solutions in which it was immersed.Conclusively, the most important conclusions are accentuated after the analysis, continued with some future research directions.Conclusions highlight the paper presented results, results that are seen as an opportunity for future studies.

Environment perception using dynamic polylines and particle based occupancy grids

Authors: Sergiu Nedevschi

Modeling and tracking dynamic entities in the driving environment is a complex task, as one has to accommodate multiple types of scenarios. Extraction of dynamic properties of obstacles becomes difficult when the measurement sensors do not provide speed directly. The dynamic polyline representation of obstacles is a compromise between the rigid model-based cuboid representation and the model-free representation of occupancy grids. This paper presents a polyline-based obstacle extraction system, based on a particle-based occupancy grid generated by processing of stereovision data. We have developed a real-time flexible method for occupancy grid modeling and representation using particles that move from cell to cell and are created and destroyed based on measurements derived from stereovision. The results of the occupancy grid processing are subjected to the Border Scanner algorithm, which extracts polylines from occupied cells, by taking into consideration only the most relevant information of the grid. The resulting system is able to extract individual objects from the occupancy grid, model them as polylines, and estimate their speed without making assumptions about their shape or size.

Intersection safety using lidar and stereo vision sensors

Authors: Sergiu Nedevschi

In this paper, we describe our approach for intersection safety developed in the scope of the European project INTERSAFE-2. A complete solution for the safety problem including the tasks of perception and risk assessment using on-board lidar and stereo-vision sensors will be presented and interesting results are shown.

An Open Source Modelica Graphic Editor Integrated with Electronic Notebooks and Interactive Simulation

Authors: Adrian Pop

This paper describes the first open source Modelica graphic editor which is integrated with interactive electronic notebooks and online interactive simulation.The work is motivated by the need for easy-to-use graphic editing of Modelica models using OpenModelica, as well as needs in teaching where the student should be able to interactively modify and simulate models in an electronic book.Models can be both textual and graphical.The interactive online simulation makes the simulation respond in real-time to model changes, which is useful in a number of contexts including immediate feedback to students.

Towards a Benchmark Suite for Modelica Compilers: Large Models

Authors: Adrian Pop

The paper presents a contribution to a Modelica benchmark suite.Basic ideas for a tool independent benchmark suite based on Python scripting along with models for testing the performance of Modelica compilers regarding large systems of equation are given.The automation of running the benchmark suite is demonstrated followed by a selection of benchmark results to determine the current limits of Modelica tools and how they scale for an increasing number of equations.

Bootstrapping a Modelica Compiler aiming at Modelica 4

Authors: Adrian Pop

What does it mean to bootstrap a compiler, and why do it?This paper reports on the first bootstrapping (i.e., a compiler can compile itself) of a full-scale EOO (Equation-based Object-Oriented) modeling language such as Modelica.The Modelica language has been modeled/implemented in the OpenModelica compiler (OMC) using an extended version of Modelica called MetaModelica.OMC models the MetaModelica language and is now compiling itself with good performance.Benefits include a more extensible maintainable compiler, also making it easier to add functionality such as debugging support.This work is in line with the recently started Modelica 4 design effort which includes moving implementation of language features from the compiler to a Modelica Core library, allowing compilers to become smaller while increasing correctness and portability.A number of language constructs discussed for Modelica 4 are already supported in some form by the bootstrapped compiler.Future work includes adapting language constructs according to the Modelica 4 design effort and extracting and restructuring parts of the Modelica implementation from the OMC compiler to instead reside in a Modelica Core library, making the compiler smaller and more extensible.

Towards a Modular and Accessible Modelica Compiler Backend

Authors: Adrian Pop

Modelica is well suited for modelling complex physical systems due to the acausal description it is using.The causalisation of the model is carried out prior to each simulation.A significant part of the causalisation process is the symbolic manipulation and optimisation of the model.Despite the growing interest in Modelica, the capabilities of symbolic manipulation and optimisation are not fully utilized.This paper presents an approach to increase the customisability, access, and reuse of symbolic optimisation by a more modular and flexible design concept.An overview of the common symbolic manipulation and optimisation algorithms of a typical Modelica compiler is presented as well as a general modular design concept for a Modelica compiler backend.The modularisation concept will be implemented in a future version of the OpenModelica compiler.

DrControl – An Interactive Course Material for Teaching Control Engineering

Authors: Adrian Pop

In this paper we present an interactive course material called DrControl for teaching control theory concepts mixed together with exercises and example models in Modelica.The active electronic notebook, OMNotebook, is the basis for the course material.This can be an alternative or complement compared to the traditional teaching method with lecturing and reading textbooks.Experience shows that using such an electronic book will lead to more engagement from the students.OMNotebook can contain interactive technical computations and text, as well as graphics.Hence it is a suitable tool for teaching, experimentation, simulation, scripting, model documentation, storage, etc.

Towards Modelica 4 Meta-Programming and Language Modeling with MetaModelica 2.0

Authors: Adrian Pop

This report gives a language definition and tutorial on how to model languages using MetaModelica 2.0 – an extension of Modelica 3.2 designed for efficient language modeling. Starting from an extremely simple language, a series of small languages are modeled by gradually adding features. Both interpretive and translational language semantics are modeled. Exercises with solutions are given. The approach of allowing the modeling language to model language semantics in principle allows the definition of language semantics in libraries, which could be used to reverse the current trend of model compilers becoming very large and complex. MetaModelica 2.0 builds on MetaModelica 1.0 which was the first Modelica language version that supports language modeling, and has been in extensive use since 2005, primarily in the development of the OpenModelica compiler. The following version of MetaModelica, called MetaModelica 2.0, is described in this report. It is easier to use since it also supports the standard Modelica 3 language features as well as additional modeling features for expressiveness and conciseness. It is implemented within the OpenModelica compiler itself. This means that the OpenModelica compiler supporting MetaModelica 2.0 is bootstrapped, i.e., it compiles itself. This work is strongly connected to the Modelica 4 effort announced by Modelica Association in September 2010, which includes moving language functionality into library packages to achieve more extensible and modular Modelica model compilers. The MetaModelica language features contribute to realizing that goal. The language features have been proven in large-scale usage in the packages within the OpenModelica compiler. However, much work still remains in improving the modularity and interface properties that are expected by library packages.

Meta-Programming and Language Modeling with MetaModelica 1.0

Authors: Adrian Pop

This report gives a language definition and tutorial on how to model languages using MetaModelica 1.0 – an extended subset of Modelica designed for efficient language modeling. Starting from an extremely simple language, a series of small languages are modeled by gradually adding features. Both interpretive and translational language semantics are modeled. Exercises with solutions are given. The approach of allowing the modeling language to model language semantics in principle allows the definition of language semantics in libraries, which could be used to reverse the current trend of model compilers becoming very large and complex. MetaModelica 1.0 is the first Modelica language version that supports language modeling, and has been in extensive use since 2005, primarily in the development of the OpenModelica compiler. MetaModelica 1.0 is strongly related to the RML specification language for Natural Semantics/Structural Operational Semantics, and is implemented using the RML compiler kernel but with a new compiler frontend. Thus, it lacks many standard language features in Modelica and requires a strictly functional modeling style. The next version of MetaModelica, becoming available during the spring 2011, is implemented within the standard OpenModelica compiler. Therefore it also supports the standard Modelica 3 language features as well as additional features for expressiveness and conciseness.

Reasoning with Forest Logic Programs and f-hybrid knowledge bases

Authors: Cristina Feier

Abstract Open Answer Set Programming (OASP) is an undecidable framework for integrating ontologies and rules. Although several decidable fragments of OASP have been identified, few reasoning procedures exist. In this paper, we provide a sound, complete, and terminating algorithm for satisfiability checking w.r.t. Forest Logic Programs (FoLPs), a fragment of OASP where rules have a tree shape and allow for inequality atoms and constants. The algorithm establishes a decidability result for FoLPs. Although believed to be decidable, so far only the decidability for two small subsets of FoLPs, local FoLPs and acyclic FoLPs, has been shown. We further introduce f-hybrid knowledge bases, a hybrid framework where knowledge bases and FoLPs coexist, and we show that reasoning with such knowledge bases can be reduced to reasoning with FoLPs only. We note that f-hybrid knowledge bases do not require the usual (weakly) DL-safety of the rule component, thus providing a genuine alternative approach to current integration approaches of ontologies and rules.

Approximate reinforcement learning: An overview

Authors: Lucian Bușoniu

Reinforcement learning (RL) allows agents to learn how to optimally interact with complex environments. Fueled by recent advances in approximation-based algorithms, RL has obtained impressive successes in robotics, artificial intelligence, control, operations research, etc. However, the scarcity of survey papers about approximate RL makes it difficult for newcomers to grasp this intricate field. With the present overview, we take a step toward alleviating this situation. We review methods for approximate RL, starting from their dynamic programming roots and organizing them into three major classes: approximate value iteration, policy iteration, and policy search. Each class is subdivided into representative categories, highlighting among others offline and online algorithms, policy gradient methods, and simulation-based techniques. We also compare the different categories of methods, and outline possible ways to enhance the reviewed algorithms.

Optimistic planning for sparsely stochastic systems

Authors: Lucian Bușoniu

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et a ̀ la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Least-squares methods for policy iteration

Authors: Lucian Bușoniu

Influence of Expert-Dependent Variability over the Performance of Noninvasive Fibrosis Assessment in Patients with Chronic Hepatitis C by Means of Texture Analysis

Authors: Sergiu Nedevschi

Texture analysis is viewed as a method to enhance the diagnosis power of classical B-mode ultrasound image. The present paper aims to evaluate and eliminate the dependence between the human expert and the performance of such a texture analysis system in predicting the cirrhosis in chronic hepatitis C patients. 125 consecutive chronic hepatitis C patients were included in this study. Ultrasound images were acquired from each patient and four human experts established regions of interest. Textural analysis tool was evaluated. The performance of this approach depends highly on the human expert that establishes the regions of interest (P < 0.05). The novel algorithm that automatically establishes regions of interest can be compared with a trained radiologist. In classical form met in the literature, the noninvasive diagnosis through texture analysis has limited utility in clinical practice. The automatic ROI establishment tool is very useful in eliminating the expert-dependent variability.

An efficient segmentation method for ultrasound images based on a semi-supervised approach and patch-based features

Authors: Sergiu Nedevschi

Segmenting ultrasound images is a challenging problem where standard unsupervised segmentation methods such as the well-known Chan-Vese method fail. We propose in this paper an efficient segmentation method for this class of images. Our proposed algorithm is based on a semi-supervised approach (user labels) and the use of image patches as data features. We also consider the Pearson distance between patches, which has been shown to be robust w.r.t speckle noise present in ultrasound images. Our results on phantom and clinical data show a very high similarity agreement with the ground truth provided by a medical expert.

Computer-Assisted Measurement of Sonoelastographic Phantom Images: Is It Reliable for Future Clinical Use?

Authors: Sergiu Nedevschi

HUMAN EXPOSURE TO POWER FREQUENCY ELECTRIC FIELD INSIDE VERY HIGH VOLTAGE SUBSTATIONS

Authors: Călin Munteanu, Vasile Topa

This paper outlines the most recent developments regarding the electric field distribution analysis inside very high voltage substations.The study was performed in order to develop and validate a powerful computation tool able to determine and predict the substation zones where the occupational exposure limits exceeds the admissible values imposed by the legislation.The study cases proposed in the paper are taken from the Romanian national power grid but the conclusions and the mitigation ideas outlined can be directly extended to substations belonging to other national or local power grids.In the first part of the paper the principles and methods proposed by the authors for the electric field computation in the vicinity of the power devices composing a very high voltage substation are outlined.The validation of the computation module implemented was done by comparisons with detailed experimental measurements results performed inside a new rehabilitated 400 kV substation.The evaluation of the results outlines a good agreement between the computed and measured electric field values inside the analyzed substation.This fact motivates the usage of the computation module developed in the design of the future high voltage substations in the light of fulfilling the actual regulations regarding the human exposure to power electric fields.In the second part of the paper scenarios to reduce the field values inside the future designed or rehabilitated substations are proposed and examined.The final conclusions end the paper.

COMPUTATION OF THE POTENTIAL INDUCED ON THE FLUID TRANSPORT PIPELINES BY OVERHEAD HIGH VOLTAGE LINES

Authors: Călin Munteanu, Vasile Topa

The paper focuses on the evaluation of the hazardous induced potentials on metallic pipelines that convey fluids such as gas, petroleum, oil and shares the same right-of-way corridors with high voltage (HV) overhead power lines (OHL).For this purpose a dedicated computation software package was developed.The 3D numerical analysis of the AC electromagnetic interferences generated by the HV lines working on normal or fault conditions to the neighbor pipeline networks is based on a coupled Finite Elements (FEM) -Boundary Elements (BEM) model.In the first part of the paper the theoretical background is outlined and the problem governing equations are presented.Details about the 1D -FEM model developed for the interior of the pipeline and the 3D -BEM model set-up for the pipeline external domain are emphasized.In the second part of the paper numerical computation examples are presented.The effect of a fault in the HV line circuit on the shape and magnitude of the hazardous induced potential distribution along the victim pipelines is detailed and discussed.The final conclusions end the paper.

Argumentative Support for Structured HACCP Plans

Authors: Adrian Petru Groza

This research presents an argumentation based decision support system for implementing the Hazard Analysis at Critical Control Points (HACCP) standard in food industry. Our analysis starts by identifying the adequate technical instrumentation needed for supporting different aspects of the HAACP system. An integrated architectural solution is presented. The framework is built around concept maps and it exploits the integration of ontologies with argumentation theory by using the Argument Interchange Format ontology.

Using Semantic Wikis for Structured Argument in Medical Domain

Authors: Adrian Petru Groza

This research applies ideas from argumentation theory in the context of semantic wikis, aiming to provide support for structured-large scale argumentation between human agents. The implemented prototype is exemplified by modelling the MMR vaccine controversy.

Generalization of an active electronic notebook for teaching multiple programming languages

Authors: Adrian Pop

In this paper we present a generalization of the active electronic notebook, OMNotebook, for handling multiple programming languages for educational purposes. OMNotebook can be an alternative or complementary tool to the traditional teaching method with lecturing and reading textbooks. Experience shows that using such an electronic book will lead to more engagement from the students. OMNotebook can contain technical computations and text, as well as graphics. Hence it is a suitable tool for teaching, experimentation, simulation, scripting, model documentation, storage, etc. OMNotebook is part of the open source platform OpenModelica. It is already used for the course material DrModelica in teaching the Modelica language but can easily be adapted to other programming languages which is also shown in this paper. The notebook can also be adapted to other areas, such as physics, chemistry, biology, biomechanics etc., where phenomena can be illustrated by dynamic simulations within the notebook. The idea behind this paper is to show that by using a standardized interface the notebook can be extended to any computer language, i.e., being language independent. This is shown in the form of an implementation and adaptation of the notebook to support the Scheme language.

The Romanian speech synthesis (RSS) corpus: Building a high quality HMM-based speech synthesis system using a high sampling rate

Authors: Adriana Stan

Middleware for Smart Environments Management

Authors: Tudor Cioara, Anghel Ionuț, Mihaela Dînșoreanu

This paper introduces a self-configuring middleware that manages the processes of context information acquisition and representation from smart closed environments, targeting the development of context aware applications. The environment context information is modeled using three sets: context resources, context actors and context policies. The context model artifacts are generated and administrated at run time by a management infrastructure based on intelligent software agents. The self-configuring property is enforced by monitoring the closed environment in order to detect variations or conditions for which the context model artifacts must be updated. The middleware was tested and validated within the premises of our Distributed Systems Research Laboratory smart environment.

An Optimization for Reasoning with Forest Logic Programs

Authors: Cristina Feier

Open Answer Set Programming (OASP) is an attractive framework for integrating ontologies and rules. In general OASP is undecidable. In previous work we provided a tableau-based algorithm for satisfiability checking w.r.t. forest logic programs, a decidable fragment of OASP, which has the forest model property. In this paper we introduce an optimized version of that algorithm achieved by means of a knowledge compilation technique. So-called unit completion structures, which are possible building blocks of a forest model, in the form of trees of depth 1, are computed in an initial step of the algorithm. Repeated computations are avoided by using these structures in a pattern-matching style when constructing a model. Furthermore we identify and discard redundant unit completion structures: a structure is redundant if there is another structure which can always replace the original structure in a forest model.

Cross-Entropy Optimization of Control Policies With Adaptive Basis Functions

Authors: Lucian Bușoniu

This paper introduces an algorithm for direct search of control policies in continuous-state discrete-action Markov decision processes. The algorithm looks for the best closed-loop policy that can be represented using a given number of basis functions (BFs), where a discrete action is assigned to each BF. The type of the BFs and their number are specified in advance and determine the complexity of the representation. Considerable flexibility is achieved by optimizing the locations and shapes of the BFs, together with the action assignments. The optimization is carried out with the cross-entropy method and evaluates the policies by their empirical return from a representative set of initial states. The return for each representative state is estimated using Monte Carlo simulations. The resulting algorithm for cross-entropy policy search with adaptive BFs is extensively evaluated in problems with two to six state variables, for which it reliably obtains good policies with only a small number of BFs. In these experiments, cross-entropy policy search requires vastly fewer BFs than value-function techniques with equidistant BFs, and outperforms policy search with a competing optimization algorithm called DIRECT.

Online least-squares policy iteration for reinforcement learning control

Authors: Lucian Bușoniu

Reinforcement learning is a promising paradigm for learning optimal control. We consider policy iteration (PI) algorithms for reinforcement learning, which iteratively evaluate and improve control policies. State-of-the-art, least-squares techniques for policy evaluation are sample-efficient and have relaxed convergence requirements. However, they are typically used in offline PI, whereas a central goal of reinforcement learning is to develop online algorithms. Therefore, we propose an online PI algorithm that evaluates policies with the so-called least-squares temporal difference for Q-functions (LSTD-Q). The crucial difference between this online least-squares policy iteration (LSPI) algorithm and its offline counterpart is that, in the online case, policy improvements must be performed once every few state transitions, using only an incomplete evaluation of the current policy. In an extensive experimental evaluation, online LSPI is found to work well for a wide range of its parameters, and to learn successfully in a real-time example. Online LSPI also compares favorably with offline LSPI and with a different flavor of online PI, which instead of LSTD-Q employs another least-squares method for policy evaluation.

Approximate dynamic programming with a fuzzy parameterization

Authors: Lucian Bușoniu

Using prior knowledge to accelerate online least-squares policy iteration

Authors: Lucian Bușoniu

Reinforcement learning (RL) is a promising paradigm for learning optimal control. Although RL is generally envisioned as working without any prior knowledge about the system, such knowledge is often available and can be exploited to great advantage. In this paper, we consider prior knowledge about the monotonicity of the control policy with respect to the system states, and we introduce an approach that exploits this type of prior knowledge to accelerate a state-of-the-art RL algorithm called online least-squares policy iteration (LSPI). Monotonic policies are appropriate for important classes of systems appearing in control applications. LSPI is a data-efficient RL algorithm that we previously extended to online learning, but that did not provide until now a way to use prior knowledge about the policy. In an empirical evaluation, online LSPI with prior knowledge learns much faster and more reliably than the original online LSPI.

Exploiting policy knowledge in online least-squares policy iteration: An empirical study

Authors: Lucian Bușoniu

peer reviewed

Reinforcement Learning and Dynamic Programming Using Function Approximators

Authors: Lucian Bușoniu

From household appliances to applications in robotics, engineered systems involving complex dynamics can only be as effective as the algorithms that control them. While Dynamic Programming (DP) has provided researchers with a way to optimally solve decision and control problems involving complex dynamic systems, its practical value was limited by algorithms that lacked the capacity to scale up to realistic problems. However, in recent years, dramatic developments in Reinforcement Learning (RL), the model-free counterpart of DP, changed our understanding of what is possible. Those developments led to the creation of reliable methods that can be applied even when a mathematical model of the system is unavailable, allowing researchers to solve challenging control problems in engineering, as well as in a variety of other disciplines, including economics, medicine, and artificial intelligence. Reinforcement Learning and Dynamic Programming Using Function Approximators provides a comprehensive and unparalleled exploration of the field of RL and DP. With a focus on continuous-variable problems, this seminal text details essential developments that have substantially altered the field over the past decade. In its pages, pioneering experts provide a concise introduction to classical RL and DP, followed by an extensive presentation of the state-of-the-art and novel methods in RL and DP with approximation. Combining algorithm development with theoretical guarantees, they elaborate on their work with illustrative examples and insightful comparisons. Three individual chapters are dedicated to representative algorithms from each of the major classes of techniques: value iteration, policy iteration, and policy search. The features and performance of these algorithms are highlighted in extensive experimental studies on a range of control applications. The recent development of applications involving complex systems has led to a surge of interest in RL and DP methods and the subsequent need for a quality resource on the subject. For graduate students and others new to the field, this book offers a thorough introduction to both the basics and emerging methods. And for those researchers and practitioners working in the fields of optimal and adaptive control, machine learning, artificial intelligence, and operations research, this resource offers a combination of practical algorithms, theoretical analysis, and comprehensive examples that they will be able to adapt and apply to their own work. Access the authors' website at www.dcsc.tudelft.nl/rlbook/ for additional material, including computer code used in the studies and information concerning new developments.

RELIABLE LOCALIZATION AND MAP BUILDING BASED ON VISUAL ODOMETRY AND EGO MOTION MODEL IN DYNAMIC ENVIRONMENT

Authors: Sergiu Nedevschi

Assessment of Sonoelastography as Diagnosis Tool of Inflammatory Myopathies

Authors: Sergiu Nedevschi

Background: Inflammatory myopathies represent a special group of pathology.Establishing the correct diagnosis in the early phase and a better follow-up are the main objective for improving the life quality of these patients.Objective: The aim of this research was to assess the usefulness of sonoelastography in diagnosis of inflammatory myopathies.Material and Method: A prospective longitudinal study with a control group was carried out in the Radio-Imaging Department, Cluj-Napoca District University Hospital, Romania, from May 2007 to July 2010.Measurements: Image Processing 2.8 software was used to analyze the elastographic images in correlation with clinical and biochemical data was used.Results: Analysis of Receiver Operating Characteristic curves statistics for elastography parameters showed significant values for the following parameters: average green color, average blue color, average intensity of color, dispersion of red, dispersion of green, and dispersion of hue.Conclusions: Our study highlighted the utility of some elastography parameters in the diagnosis of inflammatory myopathies; the parameter proved to be statistically different in case group compared to control.Anyway, standardization of sonoelastography and larger studies are required to ensure accurate diagnosis, reproductibility and reliability.

Modelling Imprecise Arguments in Description Logic

Authors: Adrian Petru Groza

Real arguments are a mixture of fuzzy linguistic variables and ontological knowledge. This paper focuses on modelling imprecise arguments in order to obtain a better interleaving of human and software agents argumentation, which might be proved useful for extending the number of real life argumentative-based applications. We propose Fuzzy Description Logic as the adequate technical instrumentation for filling the gap between human arguments and software agents arguments. A proof of concept scenario has been tested with the fuzzyDL reasoner.

Towards Unified System Modeling and Simulation with ModelicaML: Modeling of Executable Behavior Using Graphical Notations

Authors: Adrian Pop

This paper is a further step towards application of the Model-Based Systems Engineering (MBSE) paradigm, using standardized, graphical, and executable system modeling languages.It presents further development of Modelica graphical Modeling Language (Modeli-caML), a UML Profile for Modelica, which enables an integrated modeling and simulation of system requirements and design (for systems including both hardware and software).This approach combines the power of the OMG UML/SysML standardized graphical notation for system and software modeling, and the modeling and simulation power of Modelica.It facilitates the creation of executable system-specification and analysis models that can simulate time-discrete (or eventbased) and time-continuous system behavior.

Towards a Text Generation Template Language for Modelica

Authors: Adrian Pop

The uses, needs, and requirements of a text generation template language for Modelica are discussed.A template language may allow more concise and readable programming of the generation of textual models, program code, or documents, from a structured model representation such as abstract syntax trees (AST).Applications can be found in generating simulation code in other programming languages from models, generation of specialized models for various applications, generation of documentation, web pages, etc.We present several template language designs and some usage examples, both C code generation and Modelica model generation.Implementation is done in the OpenModelica environment.Two designs are currently operational.

OpenModelica-Python Interoperability Applied to Monte Carlo Simulation

Authors: Adrian Pop

During recent years the interest in computer aided modeling and simulation of complex multi-domain systems have increased significantly. The evolution of the declarative equation-based object-oriented (EOO) modeling language Modelica is a prime example of such a trend. The inherent acausal modeling possibilities and the hybrid modeling capabilities are two features that make Modelica superior. The Modelica community (Modelica 2009 [1], Fritzson 2004 [2]) has achieved considerable success through the development of libraries for a number of different technical domains. OpenModelica is an open source platform for modeling and simulation of complex physical systems based on Modelica (OpenModelica 2009 [3]). It is intended for both commercial and academic use. The development is supported by the non-profit organization Open Source Modelica Consortium. The most important parts of the platform are: OpenModelica Compiler (OMC), the interactive session handler (OMShell), and DrModelica with the electronic notebook, and an Eclipse plug-in MDT for large-scale library and compiler development [4]. For interoperability with other tools and languages Modelica models can be represented in XML (The Extensible Markup Language), is in ModelicaXML (Pop 2003[5]), and a general Java-Modelica interface [7]. Furthermore, within this platform the MetaModelica extension of the Modelica language is available for modeling the meaning of languages, e.g. Modelica, and mechanisms for model transformations. Often in a design process there is a need to make a model parameter sweep in order to find an optimal solution to a design problem, or as an improvement iteration of an existing design. For example, which design parameter values will give the minimum energy consumption in a car model. This often requires the interoperability of several tools. The aim of this paper is to illustrate the capabilities of the OpenModelica platform in this aspect. The parameter sweep is performed as a Monte Carlo simulation of a Modelica model (Kalldahl 2007 [6]). The Python language is used to set up the simulation parameters, the uncertainty distributions, as well as calling the OpenModelica compiler and simulator with the updated parameters in each step. The model variable of interest in the results can either be further analyzed in Python or exported to a Matlab file for further analyses.

A Generic Context Model Enhanced with Self-Configuring Features

Authors: Tudor Cioara, Anghel Ionuț, Mihaela Dînșoreanu

A SELF-CONFIGURING MIDDLEWARE FOR MANAGING CONTEXT AWARENESS

Authors: Tudor Cioara, Anghel Ionuț, Mihaela Dînșoreanu

Policy search with cross-entropy optimization of basis functions

Authors: Lucian Bușoniu

This paper introduces a novel algorithm for approximate policy search in continuous-state, discrete-action Markov decision processes (MDPs). Previous policy search approaches have typically used ad-hoc parameterizations developed for specific MDPs. In contrast, the novel algorithm employs a flexible policy parameterization, suitable for solving general discrete-action MDPs. The algorithm looks for the best closed-loop policy that can be represented using a given number of basis functions, where a discrete action is assigned to each basis function. The locations and shapes of the basis functions are optimized, together with the action assignments. This allows a large class of policies to be represented. The optimization is carried out with the cross-entropy method and evaluates the policies by their empirical return from a representative set of initial states. We report simulation experiments in which the algorithm reliably obtains good policies with only a small number of basis functions, albeit at sizable computational costs.

The Case of the Occasionally Cheap Computer: Low-cost Devices and Classrooms in the Developing Regions

Authors: Sergiu Nedevschi

The quest for the low-cost computer has been one of the most significant pursuits of ICTD since the 1990s. This article examines the experiences of low-cost computing projects in developing regions and looks at some of the common entrepreneurial and technical problems faced by past and current initiatives. Focusing specifically on the domain of education, we look at the quest for low-cost devices and consider their economic and socio-cultural appropriateness to the typical classroom in the developing world. Using field studies and interviews conducted in rural Indian classrooms, we show that shared rather than single-user devices constitute a more realistic and sustainable approach for low-cost computing projects targeting children’s education.

REAL TIME OBJECT DELIMITERS EXTRACTION FOR ENVIRONMENT REPRESENTATION IN DRIVING SCENARIOS

Authors: Sergiu Nedevschi

The environment representation is essential for driving assistance. However, the performances achieved in complex environments are still unsatisfactory regarding the accuracy, confidence and real time capabilities. This paper presents a real-time 2.5D environment representation model, for driving scenarios, based on object delimiters extraction from a 3D occupancy grid obtained from dense stereo. We propose two approaches to extract the polyline delimiters: an improved contour tracing called 3A Tracing and a polyline extraction method through the occupancy grid radial scanning. The advantages and drawbacks for each of these methods have been discussed.

Stereovision-Based Sensor for Intersection Assistance

Authors: Sergiu Nedevschi, Radu Gabriel Dănescu, Tiberiu Marița

Hot data centers vs. cool peers

Authors: Sergiu Nedevschi

Dramatic success in cancer immunotherapy has been achieved over the last decade with the introduction of checkpoint inhibitors, leading to response rates higher than with chemotherapy in certain cancer types. These responses are often restricted to cancers that have a high mutational burden and show pre-existing T-cell infiltrates. Despite extensive efforts, therapeutic vaccines have been mostly unsuccessful in the clinic. With the introduction of next generation sequencing, the identification of individual mutations is possible, enabling the production of personalized cancer vaccines. Combining immune check point inhibitors to overcome the immunosuppressive microenvironment and personalized cancer vaccines for directing the host immune system against the chosen antigens might be a promising treatment strategy.

Integrated Model-Driven Development Environments for Equation-Based Object-Oriented Languages

Authors: Adrian Pop

Integrated development environments are essential for efficient realization of complex industrial products, typically consisting of both software and hardware components. Powerful equation-based object-oriented (EOO) languages such as Modelica are successfully used for modeling and virtual prototyping increasingly complex physical systems and components, whereas software modeling approaches like UML, especially in the form of domain specific language subsets, are increasingly used for software systems modeling. A research hypothesis investigated to some extent in this thesis is if EOO languages can be successfully generalized also to support software modeling, thus addressing whole product modeling, and if integrated environments for such a generalized EOO language tool support can be created and effectively used on real-sized applications. However, creating advanced development environments is still a resource-consuming error-prone process that is largely manual. One rather successful approach is to have a general framework kernel, and use meta-modeling and meta-programming techniques to provide tool support for specific languages. Thus, the main goal of this research is the development of a meta-modeling approach and its associated meta-programming methods for the synthesis of model-driven product development environments that includes support for modeling and simulation. Such environments include components like model editors, compilers, debuggers and simulators. This thesis presents several contributions towards this vision in the context of EOO languages, primarily the Modelica language. Existing state-of-the art tools supporting EOO languages typically do not satisfy all user requirements with regards to analysis, management, querying, transformation, and configuration of models. Moreover, tools such as model-compilers tend to become large and monolithic. If instead it would be possible to model desired tool extensions with meta-modeling and meta-programming, within the application models themselves, the kernel tool could be made smaller, and better extensibility, modularity and flexibility could be achieved. We argue that such user requirements could be satisfied if the equation-based object-oriented languages are extended with meta-modeling and meta-programming. This thesis presents a new language that unifies EOO languages with term pattern matching and transformation typically found in functional and logic programming languages. The development, implementation, and performance of the unified language are also presented. The increased ease of use, the high abstraction, and the expressivity of the unified language are very attractive properties. However, these properties come with the drawback that programming and modeling errors are often hard to find. To overcome these issues, several methods and integrated frameworks for run-time debugging of the unified language have been designed, analyzed, implemented, and evaluated on non-trivial industrial applications. To fully support development using the unified language, an integrated model-driven development environment based on the Eclipse platform is proposed, designed, implemented, and used extensively. The development environment integrates advanced textual modeling, code browsing, debugging, etc. Graphical modeling is also supported by the development environment based on a proposed ModelicaML Modelica/UML/SysML profile. Finally, serialization, composition, and transformation operations on models are investigated.

Client Provider Collaboration for Service Bundling

Authors: Anca Nicoleta Mărginean

The key requirement for a service industry organization to reach competitive advantages through product diversification is the existence of a well defined method for building service bundles. Based on the idea that the quality of a service or its value is given by the difference between expectations and perceptions, we draw the main components of a frame that aims to support the client and the provider agent in an active collaboration meant to co-create service bundles. Following e3-value model, we structure the supporting knowledge around the relation between needs and satisfying services. We deal with different perspectives about quality through an ontological extension of Value Based Argumentation. The dialog between the client and the provider takes the form of a persuasion whose dynamic object is the current best configuration. Our approach for building service packages is a demand driven approach, allowing progressive disclosure of private knowledge.

A Comprehensive Survey of Multiagent Reinforcement Learning

Authors: Lucian Bușoniu

Multiagent systems are rapidly finding applications in a variety of domains, including robotics, distributed control, telecommunications, and economics. The complexity of many tasks arising in these domains makes them difficult to solve with preprogrammed agent behaviors. The agents must, instead, discover a solution on their own, using learning. A significant part of the research on multiagent learning concerns reinforcement learning techniques. This paper provides a comprehensive survey of multiagent reinforcement learning (MARL). A central issue in the field is the formal statement of the multiagent learning goal. Different viewpoints on this issue have led to the proposal of many different goals, among which two focal points can be distinguished: stability of the agents' learning dynamics, and adaptation to the changing behavior of the other agents. The MARL algorithms described in the literature aim---either explicitly or implicitly---at one of these two goals or at a combination of both, in a fully cooperative, fully competitive, or more general setting. A representative selection of these algorithms is discussed in detail in this paper, together with the specific issues that arise in each category. Additionally, the benefits and challenges of MARL are described along with some of the problem domains where the MARL techniques have been applied. Finally, an outlook for the field is provided.

Fuzzy Partition Optimization for Approximate Fuzzy Q-iteration

Authors: Lucian Bușoniu

Consistency of fuzzy model-based reinforcement learning

Authors: Lucian Bușoniu

Reinforcement learning (RL) is a widely used paradigm for learning control. Computing exact RL solutions is generally only possible when process states and control actions take values in a small discrete set. In practice, approximate algorithms are necessary. In this paper, we propose an approximate, model-based Q-iteration algorithm that relies on a fuzzy partition of the state space, and on a discretization of the action space. Using assumptions on the continuity of the dynamics and of the reward function, we show that the resulting algorithm is consistent, i.e., that the optimal solution is obtained asymptotically as the approximation accuracy increases. An experimental study indicates that a continuous reward function is also important for a predictable improvement in performance as the approximation accuracy increases.

Can software routers scale?

Authors: Sergiu Nedevschi

Software routers can lead us from a network of special-purpose hardware routers to one of general-purpose extensible infrastructure - if, that is, they can scale to high speeds. We identify the challenges in achieving this scalability and propose a solution: a cluster-based router architecture that uses an interconnect of commodity server platforms to build software routers that are both incrementally scalable and fully programmable.

A Sensor for Urban Driving Assistance Systems Based on Dense Stereovision

Authors: Sergiu Nedevschi, Radu Gabriel Dănescu, Tiberiu Marița, Cristian Vancea

IntroductionAdvanced driving assistance systems (ADAS) form a complex multidisciplinary research field, aimed at improving traffic efficiency and safety.A realistic analysis of the requirements and of the possibilities of the traffic environment leads to the establishment of several goals for traffic assistance, to be implemented in the near future (ADASE, INVENT, PREVENT, INTERSAFE) including: highway, rural and urban assistance, intersection management, pre-crash.While there are approaches to driving safety and efficiency that focus on the conditions exterior to the vehicle (intelligent infrastructure), it is reasonable to assume that we should expect the best results from the in-vehicle systems.Traditionally, vehicle safety is mainly defined by passive safety measures.Passive safety is achieved by a highly sophisticated design and construction of the vehicle body.The occupant cell has become a more rigid structure in order to mitigate deformations.The frontal part of vehicles has been improved as well, e.g. it incorporates specially designed "soft" areas to reduce the impact in case of a collision with a pedestrian.In the recent decades a lot of improvements have been done in this field.Similarly to the passive safety systems, primitive active safety systems, such as airbags, are only useful when the crash is actually happening, without much assessment of the situation, and sometimes they are acting against the well-being of the vehicle occupants.It has become clear that the future of the safety systems is in the realm of the artificial intelligence, systems that sense, decide and act.Sensing implies a continuous, fast and reliable estimation of the surroundings.The decision component takes into account the sensorial information and assesses the situation.For instance, a pre-crash application must decide whether the situation is of no danger, whether the crash is possible or when the crash is imminent, because depending on the situation different actions are required: warning, emergency braking or deployment of irreversible measures (internal airbags for passenger protection, or inflatable hood for pedestrian protection).While warning may be annoying, and applying the brakes potentially dangerous, deploying non-reversible safety causes permanent damage to the vehicle, and therefore the decision is not to be taken lightly.However, in a pre-crash scenario it is even more damaging if the protection systems fail to act.Therefore, it is paramount that the

An Eclipse-based Integrated Environment for Developing Executable Structural Operational Semantics Specifications

Authors: Adrian Pop

The Structural Operational Semantics Development Tooling (SOSDT) Eclipse Plugin integrates the Relational Meta-Language (RML) compiler and debugger with the Eclipse Integrated Development Environment Framework. SOSDT, together with the RML compiler and debugger, provides an environment for developing and maintaining executable Structural Operational Semantics specifications, including the Natural Semantics big step variant of SOS specifications. The RML language is successfully used at our department for writing large specifications for a range of languages like Java, Modelica, Pascal, MiniML etc. The SOSDT environment includes support for browsing, code completion through menus or popups, code checking, automatic indentation, and debugging of specifications.

Guarded Hybrid Knowledge Bases

Authors: Cristina Feier

Recently, there has been a lot of interest in the integration of Description Logics and rules on the Semantic Web.We define guarded hybrid knowledge bases (or g-hybrid knowledge bases) as knowledge bases that consist of a Description Logic knowledge base and a guarded logic program, similar to the DL+log knowledge bases from (Rosati 2006). G-hybrid knowledge bases enable an integration of Description Logics and Logic Programming where, unlike in other approaches, variables in the rules of a guarded program do not need to appear in positive non-DL atoms of the body, i.e. DL atoms can act as guards as well. Decidability of satisfiability checking of g-hybrid knowledge bases is shown for the particular DL DLRO, which is close to OWL DL, by a reduction to guarded programs under the open answer set semantics. Moreover, we show 2-EXPTIME-completeness for satisfiability checking of such g-hybrid knowledge bases. Finally, we discuss advantages and disadvantages of our approach compared with DL+log knowledge bases.

Fuzzy Approximation for Convergent Model-Based Reinforcement Learning

Authors: Lucian Bușoniu

Reinforcement learning (RL) is a learning control paradigm that provides well-understood algorithms with good convergence and consistency properties. Unfortunately, these algorithms require that process states and control actions take only discrete values. Approximate solutions using fuzzy representations have been proposed in the literature for the case when the states and possibly the actions are continuous. However, the link between these mainly heuristic solutions and the larger body of work on approximate RL, including convergence results, has not been made explicit. In this paper, we propose a fuzzy approximation structure for the Q-value iteration algorithm, and show that the resulting algorithm is convergent. The proof is based on an extension of previous results in approximate RL. We then propose a modified, serial version of the algorithm that is guaranteed to converge at least as fast as the original algorithm. An illustrative simulation example is also provided.

Continuous-state reinforcement learning with fuzzy approximation

Authors: Lucian Bușoniu

International audience

Potential of CDMA450 for rural network connectivity

Authors: Sergiu Nedevschi

This article evaluates CDMA450 as a potential solution for rural data and voice connectivity. We begin by analyzing the main strengths of CDMA450, but also some of the potential limitations for rural coverage, from both a technical and an economic standpoint. We argue that CDMA450 is a promising technology, competitive in both capacity-centric urban environments and in coverage-centric rural environments. Consequently, we discuss the opportunities of providing universal coverage by adopting a business model targeting both urban and rural deployments, and utilizing urban to rural cross-subsidization. We then explore the advantages of deploying CDMA450 using a fixed wireless (WLL) model. To this end, we explore the effective range of WLL CDMA450, and the impact of using directional antennas and receive antenna diversity in real-world commercial CDMA450 deployment in Romania. If used properly, these techniques can aid us in increasing cell radii and thus lead to substantial cost benefits

Rules with Contextually Scoped Negation

Authors: Cristina Feier

REAL-TIME FPGA-BASED IMAGE RECTIFICATION SYSTEM

Authors: Sergiu Nedevschi, Cristian Vancea

Image rectification is the process of transforming stereo-images as if they were captured using a canonical stereo-system. Computationally intensive tasks, like dense stereo matching, are greatly simplified if performed on rectified images. We developed an efficient pipeline hardware machine which performs realtime image rectification. The design was implemented using VHDL, thus allowing portability on many hardware platforms. The architecture was highly optimized, both in terms of time and resources needed. To increase its flexibility, the design was described based on generics (configuration parameters), which allow reconfiguring different characteristics and behaviour, such as: image size, number of precision bits, memory cache complexity. We also analyze the performance of the implemented solution on a VirtexE600 FPGA

ROBUST CLASSIFICATION BASED ON PRIOR OF LOCAL DIFFERENCE PROBABILITY FOR THE UNMANNED GROUND VEHICLES

Authors: Sergiu Nedevschi

The OpenModelica Modeling, Simulation, and Development Environment

Authors: Adrian Pop

Modelica is a modern, strongly typed, declarative, and object-oriented language for modeling and simulation of complex systems. This paper gives a quick overview of some aspects of the OpenModelica environment – an open-source environment for modeling, simulation, and development of Modelica applications. An introduction of the objectives of the environment is given, an overview of the architecture is outlined and a number of examples are illustrated.

Composition of XML Dialects: A ModelicaXML Case Study

Authors: Adrian Pop

This paper investigates how software composition and transformation can be applied to domain specific languages used today in modeling and simulation of physical systems. More specifically, we address the composition and transformation of the Modelica language. The composition targets the ModelicaXML dialect which is the XML representation of the Modelica language. By extending the COMPOST concrete composition layer with a component model for Modelica, we provide composition and transformation of Modelica. The design of our COMPOST extension is presented togheter with examples of composition programs for Modelica.

STEREOVISION APPROACH FOR OBSTACLE DETECTION ON NON-PLANAR ROADS

Authors: Sergiu Nedevschi, Radu Gabriel Dănescu, Tiberiu Marița

Abstract: This paper presents a high accuracy stereovision system for obstacle detection and vehicle environment perception in various driving scenarios. The system detects obstacles of all types, even at high distance, outputting them as a list of cuboids having a position in 3D coordinates, size, speed and orientation. For increasing the robustness of the obstacle detection the non-planar road model is considered. The stereovision approach was considered to solve the road-obstacle separation problem. The vertical profile of the road is obtained by fitting a first order clothoid curve on the stereo detected 3D road surface points. The obtained vertical profile is used for a better road-obstacle separation process. By consequence the grouping of the 3D points above the road in relevant objects is enhanced, and the accuracy of their positioning in the driving environment is increased. 1