Search Results (1,258)

With the increasing integration of high-penetration renewable energy and emerging loads, power system operation simulation faces two major challenges, namely strong uncertainty and significant heterogeneity in the output characteristics of multiple generator types. Traditional mathematical programming methods struggle to effectively handle uncertainty while meeting real-time computational requirements. Existing deep learning approaches fail to decouple the heterogeneous output characteristics of different generator types, which limits their ability to achieve coordinated operation. To address these issues, this paper proposes a task-decoupled and multi-task synergistic LLM-MoE method for power system operation simulation. First, a feature encoder based on Residual-Gated Linear Units is constructed to perform deep filtering and efficient representation of multi-source heterogeneous data. Second, a pre-trained large language model is employed as a temporal feature extractor to enhance temporal modeling capability and cross-scenario generalization. Finally, a customized gating-controlled mixture-of-experts decoder is developed. It dynamically coordinates task-specific and shared experts, which enables unified modeling of task decoupling, cross-task information sharing, and system physical constraints. Simulation results based on a provincial-level power grid in China demonstrate that the proposed method achieves high-accuracy and high-efficiency operation simulation while ensuring physical consistency. Full article

Prompt injection, ranked first in the OWASP Top 10 for Large Language Model (LLM) applications, enables adversaries to override system instructions and exfiltrate sensitive information by crafting inputs that blur the boundary between data and control. While application-layer defenses such as PromptShield and Prompt-G have advanced, they operate in isolation from enterprise Security Operations Center (SOC) infrastructure and lack the session-level visibility required to detect multi-turn fragmented campaigns. This paper presents a hybrid detection framework that instruments a Phi-3 Mini Instruct gateway to emit structured telemetry, correlates events in Elastic SIEM using four expert-authored detection rules, and augments rule coverage with a One-Class Support Vector Machine (OCSVM) trained exclusively on 1200 benign interactions. Evaluated against 1100 prompts (900 malicious from CySecBench, 200 benign from Stanford Alpaca), the framework achieves a precision of 0.971, a recall of 0.810, and an F1-score of 0.883, and it reduces the Attack Success Rate (ASR) to 19.0% with a Mean Time to Detection (MTTD) of 2.3 s under the evaluated Phi-3 Mini configuration. The OCSVM layer accounts for 162 of 243 incremental true positives over the baseline, identifying attacks whose behavioral feature vectors deviate from the benign manifold. The framework is architected around OpenAI-compatible gateway telemetry and is therefore designed for vendor-neutral integration; however, broader validation across model families, prompt templates, and application domains is required before making general claims about cross-model performance or production-scale effectiveness. Full article

Manual counting of surgical tools, known as surgical counting, is a time-consuming and error-prone task that increases the risk of retained surgical instruments and extends operating room (OR) time. Presently, in hospitals around the world, surgical counting is often performed manually with paper or tablet checklists, often leading to delays, increased infection risk, and financial cost. RFID, barcode-based, and computer vision solutions exist but are expensive and have challenges with sterilization and signal interference. This paper presents ToolSafe, a low-cost, portable system that classifies surgical tools by their acoustic signatures when dropped into a detection box. A pilot dataset of 4004 audio samples from four tool types (n = 996, tissue forceps; n = 1005, iris scissors; n = 1006, scalpel handle; n = 997, testing needle) was collected using ToolSafe. A convolutional neural network (CNN) was evaluated using stratified five-fold cross-validation on the laboratory dataset, with a k-nearest neighbors (KNN) classifier implemented as a control model. In each fold, both models were trained on 80% of the data and tested on the remaining 20%, ensuring that all samples were used for both training and evaluation. The CNN achieved a mean (±standard deviation) classification accuracy of 99.55% (±0.19%) across the validation folds, outperforming the KNN model, which achieved a mean accuracy of 97.28% (±0.50%). The difference was statistically significant according to a paired t-test across folds (p = 0.0003), indicating CNN’s superior performance on the dataset. For a run of 100 additional samples using the Raspberry Pi-based system, spectrogram generation averaged 0.121 s (±0.025 s), CNN inference averaged 0.180 s (±0.033 s), and total end-to-end latency averaged 1.851 s (±0.253 s) per tool. This pilot study proposes a possible technological solution for surgical counting that reduces human error and enhances patient safety. ToolSafe may be subsequently improved by increasing the number of surgical tools used in the training dataset, testing under more robust OR-like environments, and comparing to other classification algorithms. Further refinement and incorporation of ToolSafe in operating room workflows have the potential to reduce patient risks from extended surgical times and retained surgical instruments. Full article

Modern power systems are undergoing a structural transformation driven by the rapid integration of renewable energy sources, distributed energy resources, electrification, and increasing operational uncertainty. These developments expose the limitations of traditional centralized energy management and rule-based automation in highly distributed, data-intensive, and dynamically coupled energy infrastructures. In response, recent advances in artificial intelligence offer new opportunities for improving prediction, coordination, and adaptive control. This paper develops a reference architecture for Autonomous Energy Systems based on the integration of generative AI, agentic AI, digital twins, and distributed cyber–physical energy infrastructures. Rather than treating forecasting, control, simulation, and market coordination as separate research tracks, the paper organizes them within a common architectural perspective. Generative AI is positioned as a source of scenario intelligence, synthetic data generation, and uncertainty-aware forecasting, while agentic AI is framed as a bounded decision layer for perception, reasoning, planning, and coordinated action under operational constraints. The paper further clarifies the distinction between agentic AI, conventional multi-agent systems, and multi-agent reinforcement learning in energy applications. Representative application domains are discussed, including self-healing power grids, autonomous energy markets, and digital twin training environments. Major open challenges are identified in relation to scalability, physical consistency, safety verification, sim-to-real transfer, cybersecurity, interoperability with legacy infrastructures, and governance. The paper concludes by outlining a research agenda for the staged and safe development of increasingly autonomous energy systems. Full article

Extensive research has been conducted to develop technologies that enable paratransit systems to operate autonomously, including advanced sensing technologies and associated software. However, there remains a gap in research addressing adaptive operational algorithms for such systems under stochastic and dynamically evolving demand. To address this gap, this study develops an imitation-learning-assisted deep reinforcement learning (DRL) approach for autonomous shuttle routing. The proposed framework integrates generative adversarial imitation learning with proximal policy optimization to enable sequential pickup and drop-off decision-making under stochastic passenger demand without centralized re-optimization. The DRL agent was trained over approximately 1.5 million training steps and evaluated across 1000 episodes with stochastic passenger generation. Its performance was benchmarked against a deterministic dial-a-ride problem (DARP) solver implemented using Google’s OR-Tools, as well as online heuristic baselines. Results indicate that while heuristic methods achieve lower average time-based performance metrics, the proposed approach is capable of learning adaptive routing policies and demonstrates consistent behavior across diverse demand realizations. These findings highlight the feasibility of learning-based routing in controlled environments and provide a foundation for extending such approaches to more complex and realistic autonomous mobility systems. Full article

Reliable short-term agroclimatic forecasting is essential for precision agriculture, irrigation planning, disease-risk assessment, and microclimatic decision support. However, field-deployed sensor systems often operate under degraded data conditions, including missing measurements, noise, temporal interruptions, and limited local computational resources. These constraints make it necessary to evaluate not only forecasting accuracy under clean data, but also model robustness under realistic sensor-data degradation. The objective of this study is to compare machine-learning models for one-step-ahead agroclimatic time-series forecasting under degraded sensor-data conditions. Using a real meteorological dataset collected by a field weather station in the Dnipro region of Ukraine, twelve regression models were evaluated: Ridge Regression, Random Forest, Extra Trees, Gradient Boosting, HistGradientBoosting, Support Vector Regression, Linear SVR, KNN, PLSRegression, ElasticNet, Lasso, and MultiTaskElasticNet. The models were tested under five controlled scenarios: baseline data, missing values, additive noise, reduced training history, and combined noise–missingness degradation. Quantitatively, Ridge Regression achieved the strongest baseline temperature-forecasting performance, with MAE = 0.318 and R² ≈ 0.98 under clean data. It also maintained R² > 0.90 when trained on only 50% of the available history. Under Gaussian noise with σ = 0.05–0.10, Ridge Regression and HistGradientBoosting maintained R² values in the range of 0.95–0.97, whereas under combined degradation with σ = 0.10 and 20% missing data, HistGradientBoosting retained R² > 0.85. These findings indicate that machine-learning models differ substantially in their sensitivity to sensor-data degradation and that robustness-oriented benchmarking is necessary before selecting models for agroclimatic forecasting systems. Full article

The field inspection and ripeness detection of tomatoes in China remain heavily dependent on manual labor, while existing robotic solutions often exhibit limited functionality, poor environmental adaptability, prohibitive hardware costs, and unstable positioning accuracy. To address these limitations, this study proposes an intelligent tomato inspection robot that seamlessly integrates real-time ripeness recognition with precise spatial localization. Built upon a Raspberry Pi 5 core controller, the robot employs a lightweight, layered modular architecture designed to flexibly navigate complex agricultural environments. A comprehensive, multi-dimensional image dataset of tomato ripeness was constructed to train a three-category detection model based on the YOLOv8n architecture. Following 413 training epochs, the model demonstrated exceptional performance, achieving an overall mAP@0.5 of 87.8% and an mAP@0.5:0.95 of 72.7% on the held-out test dataset. In field inspections, the system achieved detection precisions of 82.22% for immature tomatoes, 92.66% for half-ripened tomatoes, and 100% for fully ripe tomatoes, successfully identifying all ripe tomatoes and satisfying the practical demands of field inspection. Furthermore, the integration of an Ultra-Wideband positioning system yielded an overall Root Mean Square Error of 0.231 m, successfully confining positioning errors to within 0.24 m to fully satisfy the stringent localization demands of crop-level inspection. Field evaluations confirmed that under optimal configurations, the robot can efficiently inspect a 50-m planting row in 10 min (±1 min) and maintains a continuous operational battery life of 2 h (±10 min). The core contribution of this work is the system-level integration and optimization of technologies for greenhouse agriculture. This integrated design achieves low hardware cost and high deployment flexibility, addressing longstanding challenges of labor-intensive inspection and delayed harvesting, and delivering a practical solution for intelligent tomato plantation management. Full article

The convergence of artificial intelligence and robotic surgery is redefining the management of genitourinary cancers by enhancing diagnostic accuracy, surgical precision, and training efficiency. This narrative review explores recent advancements in artificial intelligence applications across the cancer care continuum, with a focus on prostate, kidney, and bladder malignancies. Artificial intelligence tools, particularly those based on machine learning and deep learning, have demonstrated strong performance in analyzing imaging data, segmenting tumors, predicting pathological features, and supporting clinical decision-making. Intraoperatively, artificial intelligence enables skill assessment, personalized feedback, and real-time navigation by processing data from surgical videos and robotic system sensors. Augmented reality and intraoperative modeling further enhance visualization and margin control during complex procedures. The review also discusses emerging technologies such as single-port robotic platforms, which offer advantages in confined anatomical spaces and support less invasive approaches. Additionally, the growing field of telesurgery is addressed, highlighting its feasibility for complex urologic operations across vast distances. While many of these innovations are still in early stages of clinical validation, their integration into practice has the potential to improve oncologic and functional outcomes, expand access to expert care, and foster the development of next-generation surgical strategies in urologic oncology. Full article

Radio Frequency Identification (RFID) is an emerging technology in Industry 4.0 for low-cost logistics, yet direction and orientation estimation typically requires multiple antennas, and robustness under industrial multipath fading, operator variability, and signal fragmentation has not been evaluated. To address this gap, this study proposes a single-antenna RFID system that evaluated thirteen architectures spanning unsupervised methods (clustering algorithms) and supervised methods (classical machine learning, deep learning, and hybrid architectures) on Received Signal Strength Indicator (RSSI) and phase time-series reconstructed through a pipeline of Savitzky–Golay smoothing, phase unwrapping, and cubic spline resampling to $N$ = 50–300 samples, preserving signal morphology across variable-length RFID passes. The system further incorporates a physics-informed augmentation strategy that encodes multipath fading, distance variation, and fragmentation into synthetic training samples for cross-domain generalization without hardware modification. In controlled laboratory experiments, both direction and orientation tasks achieved $>$99.5% accuracy, while direction tracking was additionally validated on an industrial shop floor under varying distances, Non-Line-of-Sight (NLoS) occlusions, and signal fragmentation. Zero-shot transfer caused accuracy to degrade to near-chance levels for several configurations, confirming a pronounced domain gap. Domain adaptation with XGBoost recovered direction accuracy to $>$97% under severe fragmentation under NLoS conditions, with an inference latency of ≈150 $\mu$s. Under domain-adapted shop floor conditions, direction accuracy exceeded the 75–92% reported in prior single-antenna laboratory studies, suggesting that physics-informed domain adaptation is a promising approach for single-antenna RFID tracking in Industrial Internet of Things (IIoT) logistics environments. Full article

The widespread adoption of Generative Artificial Intelligence (GenAI) in regulated enterprises is currently hindered by the “Static Alignment Trap”: the inability of traditional safety methods, such as Reinforcement Learning from Human Feedback (RLHF), to adapt to rapidly shifting compliance landscapes without costly retraining. This paper introduces and evaluates the Agreement Validation Interface (AVI), a modular governance architecture that functions as a deterministic middleware layer. By decoupling governance from the core inference engine, AVI implements Dynamic Bilateral Alignment (DBA), enforcing policy constraints at both the input and output stages through vector-based semantic retrieval. Adopting a Design Science Research (DSR) methodology, we validated the system against the FinanceBench financial benchmark ($N=150$ queries, three repeated runs, 450 total observations) and a proprietary Russian-language provocative content dataset developed internally at MWS AI ($N=201$ queries; not publicly available). The empirical results demonstrate that the architecture achieves an 83.2% Large Language Model (LLM)-judge compliance rate (95% confidence interval, CI: 79.4–87.1%), statistically significantly exceeding the unfiltered baseline of 63.7% ($\Delta =+19.5$ percentage points (pp), $t=4.02$, $p=0.002$). The vector-based input filter achieves perfect detection performance (Precision $=1.000$, Recall $=1.000$, F1 $=1.000$). Cross-domain validation on 201 Russian-language provocative queries confirms generalizability (Recall $=0.985$, LLM compliance among triggered queries $=0.977$). The operational Time-to-Compliance for enforcing new rules was reduced from hours (model fine-tuning) to under five seconds (vector indexing). These findings suggest that enterprise AI safety requires an architectural shift from model-centric training to system-centric control, complemented by system-prompt-level anti-inference governance. We conclude that AVI offers a scalable, cost-effective, and statistically validated framework for auditable AI compliance, independent of the underlying model provider. Full article

The increasing penetration of single-phase photovoltaic (PV) generation and electric vehicle (EV) charging has aggravated phase current asymmetry in low-voltage distribution networks. In contrast to voltage-oriented control strategies, this work focuses directly on mitigating current imbalance at the point of common coupling (PCC). A coordinated control framework based on multi-agent deep deterministic policy gradient (MADDPG) is developed to regulate distributed battery energy storage systems (BESS). The control objective is formulated in terms of the Current Unbalance Factor (IUF), derived from symmetrical component theory. A linearized DistFlow model is embedded in the learning environment to preserve physical consistency while maintaining computational tractability. Device-level constraints, including state-of-charge limits and ramp-rate bounds, are enforced through action projection, whereas network security limits are incorporated via reward penalties. Case studies on a modified residential feeder indicate that coordinated BESS control reduces the peak IUF from 2.75% to 2.50% under the studied operating condition. The maximum dominant-phase current decreases from 125 A to 110 A. The performance is close to that of centralized convex optimization while enabling decentralized real-time execution after offline training. These results suggest that multi-agent reinforcement learning can serve as a feasible alternative for phase imbalance mitigation in distribution networks with high renewable penetration. Full article

Railway information systems routinely register the displacement of trains across the network as sequences of station passages and segment traversals. This paper proposes a design-oriented framework that systematically transforms such train displacements into event logs to enable established process mining analyses. Here, design-oriented means that the event log is not assumed to be readily available, but is explicitly constructed from railway records through modelling choices grounded in operational semantics. The framework comprises: (i) an eventization pipeline that maps displacements to semantically precise events with explicit lifecycle and case notions; (ii) construction of a timetable-derived reference model representing planned control flow; and (iii) a structural comparison and variant analysis stage that identifies execution-level deviations from the timetable-derived reference and organizes them into recurrent behavioural patterns. The paper contributes design principles for mapping train displacements into process-mining events, a timetable-derived representation of expected control flow, and an empirical demonstration on real-world railway data showing how this framework supports operational process analysis. Full article

Shipboard boarding systems operating under complex sea conditions are subject to vessel motion coupling, wave induced disturbances, strong nonlinearity, and engineering constraints, which make accurate end pose tracking difficult. Existing mechanism-based approaches often suffer from modeling inaccuracies and high online computational burden, whereas purely data driven methods usually provide limited interpretability for safety critical marine applications. To address these limitations, this paper proposes a data driven predictive control method for shipboard boarding systems based on a Fuzzy Neural Broad Learning System. An interpretable Linear Regression Decision Tree is first constructed to represent the plant through state space partition and local linear approximation. On this basis, a Fuzzy Neural Broad Learning predictor is developed to capture disturbance-induced uncertainty and parameter variation with fast analytical training and incremental updating capability. The predictor is then embedded into a constrained model-predictive control framework in which actuator saturation, input rate limits, and output safety constraints are handled explicitly, and closed-loop boundedness is analyzed theoretically. Simulation results on a MATLAB R2024a-based and Simulink-based coupled platform show that, for the translational outputs of the gangway end effector, the testing root mean square error ranges from 1.33 × 10⁻³ to 1.74 × 10⁻³, with corresponding coefficients of determination ranging from 0.820 to 0.912. In comparative closed-loop simulations against proportional integral derivative control, fuzzy control, and learning-based control under identical operating conditions, the proposed method achieves the lowest integral of squared error and integral of absolute error, reaching 3.40 × 10⁻⁷ and 4.28 × 10⁻⁴, respectively. Compared with the best value among the three baseline controllers, the proposed method reduces the integral of squared error by approximately 42.6% and the integral of absolute error by approximately 34.4%. Although its maximum deviation is not the smallest among all compared controllers, it remains within the same order of magnitude as the advanced baselines. In addition, the average and maximum per-step computation times are 1.61 × 10⁻⁴ s and 3.75 × 10⁻³ s, respectively, both of which are far below the adopted sampling period of 0.05 s. These results indicate that the proposed framework improves cumulative tracking accuracy while maintaining feasible online computational performance. Full article

Manual assembly remains critical in Industry 5.0 high-mix/low-volume manufacturing, but it introduces resilience challenges due to cognitive load, training demands and frequent product changes. While AI-based supervision can mitigate errors, deploying such systems is often hindered by the cost of collecting and labelling thousands of real images for each product variant. This paper presents a Human-in-the-Loop semantic-twin pipeline that generates approximately 45,000 labelled synthetic images from a single CAD-based configuration and uses them to train an object detection model for real-time assembly supervision. Experiments on seven virtual environment configurations show that removing realistic lighting or camera motion reduces F1-score on real images from 0.87 to 0.46, confirming their critical role for synthetic-to-real transfer. A controlled laboratory study on a single bicycle chainring assembly task with 10 participants and 100 monitored cycles demonstrates the feasibility of automatic KPI extraction, with error events associated with a 25.6% increase in average cycle time (from 58.4 s to 73.3 s) under the tested conditions. Compared to manual annotation, where labelling 3000 images required approximately 4 h, the semantic-twin configuration takes around 4 to 6 h including image generation that enables rapid creation of large labelled datasets for new product variants without additional human annotation. These results provide a proof-of-concept foundation for resilient, data-efficient supervision of high-mix manual workstations, with full industrial validation across multiple products, stations and operator demographics identified as the necessary next step. Full article

Water utilities frequently perform pipeline-emptying operations for maintenance, repair, and operational management. This process involves transient flow conditions with entrapped air. It must be carefully controlled, as the expansion of air pockets can generate sub-atmospheric pressures that may lead to pipeline collapse. The mathematical modelling of emptying processes with air valves has been extensively studied in recent years; however, such approaches typically rely on complex algebraic–differential equation systems. This study advances understanding of this phenomenon by proposing a novel procedure that uses a machine learning model to approximate system behaviour while avoiding fully coupled hydraulic formulations. An experimental facility consisting of a pipeline with an internal diameter of 0.042 m and a total length of 4.6 m was used, in conjunction with a complete regulation valve manoeuvre. The system was first calibrated using experimental data and subsequently employed in Monte Carlo simulations to generate a dataset for training the machine learning model. The results demonstrate that a Rational Quadratic Gaussian Process Regression model can accurately predict the minimum sub-atmospheric pressure, achieving a coefficient of determination greater than 0.999 during validation and testing. The proposed framework is presented as a proof-of-concept and has been validated only for the specific case study analysed. While the results highlight its potential to support planning for emptying operations under varying air-admission conditions and air-pocket sizes, further validation is required before generalising to real-world water distribution systems. For practical implementation, the model must be appropriately trained for each specific installation. Full article