Search Results (9,282)

Anomaly detection is crucial in maintaining the safety, reliability, and optimal performance of complex systems across diverse domains, such as industrial manufacturing, cybersecurity, and autonomous systems. While conventional methods typically handle single data modalities, recently, there has been an increase in the application of multimodal detection in dynamic real-world environments. This paper presents a comprehensive review of recent research at the intersection of agentic artificial intelligence and large language-based multimodal anomaly detection. We systematically analyze and categorize existing studies based on the agent architecture, reasoning capabilities, tool integration, and modality scope. The main contribution of this work is a novel taxonomy that unifies agentic and multimodal anomaly detection methods, alongside benchmark datasets, evaluation methods, key challenges, and mitigation strategies. Furthermore, we identify major open issues, including data alignment, scalability, reliability, explainability, and evaluation standardization. Finally, we outline future research directions, with a particular emphasis on trustworthy autonomous agents, efficient multimodal fusion, human-in-the-loop systems, and real-world deployment in safety-critical applications. Full article

The current literature on electrical system maintenance highlights three technology domains—building information modeling (BIM), Digital Twin (DT), and extended reality (XR)—that have independently demonstrated strong potential for improving lifecycle information management, predictive analytics, and operational support. However, their convergence remains largely underexplored, particularly in electrical system maintenance. This paper provides a structured review of BIM–DT–XR convergence in electrical system lifecycle management, examining their roles across lifecycle phases and their integration through literature synthesis and cross-domain implementation evidence. BIM is analyzed as a basis for modeling and integrating facility management with electrical asset lifecycles; DT as a framework for dynamic system representation and applications in electrical and power systems; and XR as a means of visualizing and interacting with BIM-DT environments. Cross-domain implementation evidence from an industrial electrical facility and a tertiary smart-building pilot shows that BIM–DT–XR integration is technically feasible at pilot scale. However, the analysis identifies five structural integration gaps: semantic misalignment between building-oriented IFC and grid-oriented CIM ontologies; fragmented standard adoption; inconsistent data governance and naming practices; validation approaches focused on syntactic rather than dynamic model fidelity; and the separation of XR visualization from predictive DT capabilities. The implementation evidence further indicates that real-world deployment remains constrained by data quality limitations, integration complexity, cost factors, and interoperability with legacy systems. The review concludes that, despite the maturity of individual technologies, their effective application depends on advances in semantic alignment, lifecycle data governance, validation of dynamic models, and scalable integration frameworks, enabling the transition toward integrated, interoperable, and lifecycle-aware infrastructures for electrical system maintenance. Full article

Photoplethysmography (PPG) has emerged as an attractive modality for non-invasive cardiovascular monitoring due to its low cost, unobtrusive nature, and ubiquity in consumer wearable devices. Despite its potential, existing PPG-based arrhythmia detection systems remain limited in scope: (i) most target only atrial fibrillation, (ii) temporal localization of abnormal segments is rarely provided, and (iii) deep learning models lack explainability, hindering adoption in clinical workflows. We present a comprehensive and fully integrated framework for multi-class arrhythmia detection, segmentation, and explainability based on PPG waveforms, Heart Rate Variability (HRV), and structured clinical metadata. The proposed system introduces a CLIP-style contrastive learning module aligning PPG waveforms with clinical variables and rhythm-state textual descriptions using BioBERT; a multitask U-Net architecture performing 4-class classification and 1D segmentation; a Retrieval-Augmented Generation (RAG) pipeline leveraging Gemini Flash large language models to produce guideline-grounded diagnostic reports; and a real-time Streamlit-based web platform supporting inference, visualization, and database storage. The system significantly improves classification accuracy (from 86.27% to 91.19%) and segmentation Dice (from 0.5815 to 0.7167). These results demonstrate the feasibility of a robust, multimodal, and explainable PPG-based arrhythmia monitoring system for real-world applications. Full article

Disruption management is critical for airline operations, yet existing recovery models often assume deterministic disruption durations, limiting their effectiveness in real-world, uncertain environments. This paper addresses the integrated airline recovery problem under uncertain disruptions. To capture this uncertainty, delay times are modeled as fuzzy variables and a fuzzy chance-constrained programming model is developed, aimed at minimizing total recovery costs. The model is transformed into a deterministic equivalent using trapezoidal fuzzy numbers. An improved Greedy Randomized Adaptive Search Procedure (GRASP) algorithm is designed to efficiently solve the problem, balancing solution quality and computational efficiency through insert, exchange, and cancel. The local search process is enhanced by incorporating the acceptance criteria of the simulated annealing algorithm. The proposed method is validated using real-world airline data. Results show that, compared to the traditional GRASP algorithm, the improved GRASP algorithm can obtain better solutions in a shorter time; the solutions in deterministic scenarios tends to be more conservative, leading to resource waste; the proposed method can achieve airline recovery at the minimum recovery cost. Sensitivity analysis reveals that selecting an appropriate confidence level significantly influences recovery costs. This paper provides a robust framework for enhancing operational resilience and passenger satisfaction under uncertain conditions, offering practical insights for real-world application. Full article

Research on information and communication technologies (ICTs) in sustainable agriculture has largely been technocentric, focusing on effectiveness, efficiency, and adoption, with limited consideration of end-user perceptions in practice. This study addresses this gap by examining perceptions of mobile gardening apps as accessible ICT tools that may support sustainable behaviours. Based on over 180,000 user reviews from Google Play and the Apple App Store, Contextualized Topic Modeling (CTM) was used to identify key themes and interpret them within the Theory of Consumption Value (TCV) framework. This approach allows for the analysis of functional, emotional, and epistemic dimensions of user experiences based on large-scale, real-world data. The results indicate that functional aspects, such as reliability and usability, dominate app evaluation, but emotional engagement and knowledge acquisition also play a significant role. By combining a data-driven approach with a well-established behavioural framework, this study bridges the gap between technological and user perspectives. It simultaneously extends the application of the TCV to the field of ICT solutions supporting sustainable development and provides practical guidance for designing more effective gardening apps. Full article

The widespread adoption of large-scale sensor networks in privacy-sensitive and safety-critical applications has intensified the demand for secure, trustworthy, and energy-efficient learning mechanisms at the network edge. Federated learning has emerged as a promising paradigm for privacy preservation by enabling collaborative model training without sharing raw sensor data. However, most existing federated approaches inadequately address trust management, communication efficiency, and energy constraints, which are critical in real-world sensor-based systems. This paper proposes a trust-aware and energy-efficient federated learning framework specifically designed for secure sensor networks operating in resource-constrained edge environments. The proposed approach integrates lightweight trust metrics, trust-driven model aggregation, and adaptive communication scheduling to mitigate the impact of unreliable or malicious nodes while reducing unnecessary energy expenditure. By dynamically weighting client contributions based on trust and participation efficiency, the framework enhances robustness and learning stability under heterogeneous sensing conditions. Experimental results show that the proposed method maintains significantly higher accuracy under adversarial participation while reducing communication overhead and cumulative energy consumption. In particular, the framework improves model accuracy by up to 3.2% under heterogeneous conditions, reduces communication overhead by 28%, and decreases cumulative energy consumption by 31% compared with conventional federated learning approaches. Full article

Vision–language models (VLMs), such as Contrastive Language–Image Pretraining (CLIP), are increasingly deployed in real-world applications, including content moderation, misinformation detection, and fraud analysis, making their robustness to adversarial attacks a critical concern. While adversarial robustness has been widely studied in unimodal models, modality-specific vulnerabilities in multimodal models remain underexplored. In this work, we analyze CLIP by applying gradient-based adversarial attacks to its vision and language modalities, both independently and jointly, and evaluating performance on two multimodal classification benchmarks: the Facebook Hateful Memes dataset and a large-scale Suspicious Car Parts dataset. Using Fast Gradient Sign Method (FGSM) and Projected Gradient Descent (PGD) attacks along with multiple adversarial retraining strategies, we show that adversarial perturbations on the image modality consistently cause the most severe and unstable performance degradation. These results demonstrate that the vision modality is the primary vulnerability in CLIP, highlighting the need for modality-specific defense strategies that focus more on the weaker modality in multimodal systems. Full article

This research examines the emerging ecosystem of models that are developed and run across a distributed network of computers called decentralized artificial intelligence. The focus is to understand these models in the healthcare context and with a focus on their core components: technologies, governance frameworks, and real-world applications. A systematic literature review was conducted, analyzing peer-reviewed studies from PubMed, Scopus, and Web of Science to map the current landscape of the field. The primary objective was to synthesize the current research on decentralized approaches in healthcare, including core approaches like federated learning and blockchain-based AI models, as well as emerging concepts such as agentic AI blockchain-based AI models and DAOs, to comprehend their application in clinical and operational settings. The research assesses the maturity of these implementations, ranging from pilot programs to large-scale organizational settings. It also identified the key computational and technical methods and platforms used and the key benefits and challenges influencing their adoption. The findings underscore the pivotal role of the decentralized paradigm in addressing the fundamental limitations of traditional AI, including data privacy, trust, institutional silos, and regulatory complexity. Insights are also offered for healthcare providers, technology developers, researchers, and policymakers aiming to navigate and leverage decentralized AI to build more equitable, efficient, and collaborative healthcare systems. Full article

To meet the practical application requirements of underwater biomimetic robots, this paper presents the design of a flexible pectoral fin with integrated sensing and actuation capabilities, based on a “material-structure-function” integrated approach. The sensor film is embedded into the pectoral fin via an embedded cast-molding method, ensuring synchronized deformation and long-term cyclic stability. Experimental results demonstrate that the integrated pectoral fin can accurately perceive its own bending deformation and external environmental disturbances, enabling corresponding obstacle avoidance maneuvers in a manta robot prototype. This design strategy endows the manta robot with environmental adaptability for real-world applications and offers a novel paradigm for the intelligent design of other underwater equipment. Full article

Estimating remaining useful life (RUL) and predicting bearing faults based on data-driven models have become central components of modern Prognostics and Health Management (PHM) systems. Although deep learning models have demonstrated strong performance under controlled and stationary operating conditions, their reliability in real-world industrial and marine environments is limited. In practice, operating conditions, sensor properties, and degradation mechanisms evolve continuously over time, leading to non-stationary and shifting data distributions that violate the assumptions of conventional static learning approaches. To address these challenges, two research areas have gained increasing attention: Domain Adaptation (DA), which aims to mitigate distribution discrepancies across operating conditions or machines, and Continual Learning (CL), which enables models to learn sequentially while mitigating catastrophic forgetting. However, existing studies often examine these paradigms in isolation, limiting their effectiveness in long-term deployments, where domain shifts and temporal evolution coexist. This paper presents a comprehensive and systematic review of data-driven methods for bearing fault prognosis and remaining useful life (RUL) prediction under evolving data distributions, adopting the framework of Domain-Adaptive Continual Learning (DACL). By jointly examining the DA and CL methods, this review analyses how these approaches have been individually and implicitly combined to cope with non-stationarity, knowledge retention, and limited label availability in practical PHM scenarios. We categorised existing methods, highlighted their underlying assumptions and limitations, and critically assessed their applicability to long-term, real-world monitoring systems. Furthermore, key open challenges, including scalability, robustness under sequential domain shifts, uncertainty handling, and plasticity–stability trade-offs, are identified, and research directions are outlined based on the identified limitations and practical deployment requirements of the proposed method. This review aims to establish a structured and critical reference framework for understanding the role of domain-adaptive CL in data-driven prognostics, clarifying current research trends, limitations, and open challenges in evolving data distributions. Full article

Feature weighting plays a central role in medical classification by enhancing predictive accuracy, interpretability, and clinical trust through the explicit quantification of variable relevance. Despite their widespread use, existing filter-, wrapper-, and embedded-based feature weighting methods are predominantly deterministic and exhibit pronounced sensitivity to label noise and outliers, which are pervasive in real-world medical data. This often results in unstable importance estimates and unreliable clinical interpretations. In this work, we introduce a novel Bayesian feature weighting model that fundamentally departs from existing approaches by jointly integrating simplex-constrained Dirichlet priors for global feature weights, hierarchical shrinkage priors for coefficient regularization, and contamination-aware priors for explicit modeling of label noise within a single coherent probabilistic framework. Unlike conventional Bayesian feature selection or robust classification models, the proposed formulation yields globally interpretable feature weights defined on the probability simplex, while simultaneously providing full posterior uncertainty quantification and robustness to both mislabeled observations and aberrant feature values through principled influence control. Comprehensive simulation studies across diverse contamination scenarios, together with applications to multiple real-world medical datasets, demonstrate that the proposed model consistently outperforms classical and state-of-the-art baselines in terms of discrimination, probabilistic calibration, and stability of feature-importance estimates. These results highlight the practical and methodological significance of the proposed framework as a robust, uncertainty-aware, and interpretable solution for medical decision making under noisy data conditions. Full article

This review provides a comprehensive overview of the most adopted 3D surveying techniques in Cultural Heritage, offering practical guidance for the selection of appropriate methods when three-dimensional documentation of artworks is required. The analysis focuses on the most effective technologies for the 3D documentation of sites and objects of artistic value, with selection criteria primarily centred on non-invasiveness, given the uniqueness and cultural significance of the case studies, and the instrument flexibility, a crucial requirement for non-transportable items. A broad spectrum of 3D techniques is currently available for the multiscale diagnostic investigation of artworks, providing information at both macroscopic and microscopic levels. This review reports on the state of the art of such systems and evaluates the main characteristics of each technology in relation to its applicability in the heritage field. Particular attention is given to highlighting advantages and limitations, and to assessing performance in terms of resolution, gauge volume/area, acquisition time, and cost. In addition, the review discusses exemplary cases in which 3D methods are integrated with other analytical techniques to enable a more comprehensive understanding of the object under investigation. Finally, recent studies are examined to identify the most suitable approaches and the specific requirements for the digitization of real-world heritage assets. Full article

The industrial metaverse is a digital twin space that integrates the real world with virtual environments through bidirectional synchronization. It supports critical services, such as time-sensitive machine control and large-scale collaboration, which require Time-Sensitive Networking and scalable Data Distribution Services. DDS, developed by the Object Management Group, provides excellent scalability and diverse QoS policies but struggles to guarantee transmission delay and jitter for time-critical applications. TSN, based on IEEE 802.1 standards, addresses these challenges by ensuring time-criticality. However, current research lacks comprehensive integration mechanisms for DDS and TSN, particularly from the viewpoints of semantics and system framework. Additionally, there is no adaptive QoS mapping converting the abstract DDS QoS policies to the sophisticated TSN QoS parameters. This paper presents a novel DDS-over-TSN framework that incorporates three key functions to address these challenges. First, Cross-layer QoS Mapping automates correspondences between DDS and TSN parameters, deriving technical constraints from standard documentation through retrieval-augmented generation. Second, Semantic Priority Estimation extracts substantial priority levels by utilizing language model embedding vectors as high-dimensional feature extractors. Third, Adaptive Resource Allocation performs dynamic bandwidth distribution for each priority level through reinforcement learning. Simulation results reveal over 99% mapping accuracy and 97% consistency in priority extraction. The applied Deep Reinforcement Learning paradigm allocated 99% of required resources to high-priority classes and reduced resource wastage by 15% compared to conventional methods. This methodology meets industrial requirements by ensuring both deterministic real-time performance and efficient resource isolation. Full article

Urban logistics systems face growing delivery demand and complex traffic and operational constraints, which make unmanned delivery carriers, including unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs), a promising solution. Existing studies typically focus on a single delivery carrier type and rely on idealized assumptions, overlooking heterogeneous cooperation under multiple stations, multiple time windows, and real-world transport conditions. To address these gaps, we propose the Multi-Station UAV–UGV Cooperative Delivery Scheduling Problem with Temporally Discontinuous Service Availability (MSUUCDSP) to minimize the total travel and waiting time of UAVs and UGVs. To solve the problem, we propose a mixed-integer linear programming (MILP) model with a novel mathematical approach and a Hybrid Large Neighborhood Search (HLNS) algorithm. Additionally, we adopt a Hidden Markov Model (HMM)-based map-matching method and big data techniques to capture realistic operational characteristics. Computational experiments are conducted on various realistic instances under four diverse scenarios. Results show that UAV–UGV cooperation significantly improves efficiency, reducing total time cost by 17.12% compared with single-mode delivery, and they reveal substantial discrepancies between idealized assumptions and realistic scenarios. We further develop an ArcGIS-based simulation to support practical implementation. The findings provide valuable insights for decision-making and engineering applications for logistics operators. Full article

Pedestrian gender recognition plays an important role in pedestrian analysis and intelligent video applications, for example, in demographic statistics, soft biometric analysis, and context-aware person retrieval. However, it remains a challenging task owing to viewpoint variations, illumination changes, occlusions, and low image quality in real-world imagery. To address these issues, an effective adaptive decision fusion framework, termed the Decision Fusion Learning Network (DFLN), is proposed in this paper. The key novel aspect of DFLN is that it effectively explores both an appearance-centered view that emphasizes detailed texture and clothing information and a structure-centered view that captures rich contour and structural information for pedestrian gender recognition. To realize DFLN, a Parallel CNN Prediction Probability Learning Module (PCNNM) is first constructed to independently learn modality-specific probabilities from color image and edge maps. Subsequently, a learnable Decision Fusion Module (DFM) is designed to fuse the modality-specific probabilities and explore their complementary merits for realizing accurate pedestrian gender recognition. The DFM can be easily coupled with the PCNNM, forming an end-to-end decision fusion learning framework that simultaneously learns the feature representations and carries out adaptive decision fusion. Experiments on two pedestrian benchmark datasets, named PETA and PA-100K, show that DFLN achieves competitive or superior performance compared with several state-of-the-art pedestrian gender recognition methods. Extensive experimental analysis further confirms the effectiveness of the proposed decision fusion strategy and its favorable generalization ability under domain shift. Full article