Search Results (483)

Background/Objectives: Pharmacovigilance workflows rely heavily on unstructured text across diverse sources. Here, we systematically reviewed how large language models (LLMs) are being explored as support tools for adverse drug reaction (ADR) detection, extraction, triage, and documentation, highlighting their potential for precision medicine and big data-enabled safety monitoring. Methods: Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2020 guidelines, we systematically searched PubMed, Scopus, and Web of Science for studies published between January 2022 and March 2026. Ultimately, 83 empirical studies satisfied the inclusion criteria. A narrative synthesis was conducted to address methodological heterogeneity across these studies. Results: LLM applications were concentrated in constrained information-extraction and classification tasks, including signal evaluation, clinical-note extraction, social media surveillance, and literature screening. Quantitative performance varied substantially by system design: error-correction prompting yielded an F1-score of 0.921 for ADR named entity recognition, whereas retrieval-augmented generation improved data-retrieval accuracy from 8.3% to 78.3%. Most studies were retrospective, benchmark-based, or proof-of-concept evaluations. Across 581 paired pre-consensus domain judgements, observed inter-rater agreement was 90.4% and Cohen’s κ was 0.837 (95% CI 0.772–0.895). Hallucination, low specificity, prompt sensitivity, narrow datasets, and weak external validation remained common limitations. Conclusions: Current evidence supports supervised, task-specific applications of LLMs for extraction, triage, retrieval, and documentation rather than autonomous pharmacovigilance decision-making. Prospective evaluation, external validation, transparent reporting, and accountable human oversight are required before high-stakes clinical or regulatory deployment. Full article

Background/Objectives: Artificial intelligence (AI) is increasingly being evaluated for ophthalmic diagnosis, screening, and triage, yet its role in paediatric eye care remains less established than in adult ophthalmology. This systematic review aimed to synthesise evidence on AI-enabled tools for paediatric ophthalmic diagnosis, screening, triage, surveillance, and referral, with an emphasis on diagnostic performance, safety, workflow integration, equity, and implementation readiness in primary, community, and primary care-relevant settings. Methods: A PRISMA-guided systematic review was conducted using MEDLINE, Embase, Web of Science, Scopus, and IEEE Xplore from inception to 30 March 2026. Eligible studies evaluated AI or machine-learning tools for children and adolescents aged 0–18 years in relation to paediatric eye conditions. Study selection and data extraction were undertaken independently by reviewers, with disagreements resolved by consensus or third-reviewer adjudication. Methodological and reporting quality was evaluated using an author-adapted six-domain rubric informed by APPRAISE-AI. Diagnostic-accuracy studies were assessed using an author-adapted QUADAS-2 framework incorporating QUADAS-AI-informed AI-specific considerations, the prediction-model study was assessed using PROBAST+AI, and the non-randomised treatment-effect study was assessed using ROBINS-I. The public dataset descriptor was evaluated separately using an author-developed dataset-quality, representativeness, and applicability framework. Because of clinical and methodological heterogeneity, findings were synthesised thematically. Results: Twelve empirical studies and one public dataset descriptor were included, covering retinopathy of prematurity, retinoblastoma, amblyopia risk, myopia, congenital cataract, and visual-acuity assessment. AI systems frequently demonstrated promising diagnostic or screening performance, including sensitivity-first detection of treatment-requiring retinopathy of prematurity, high discrimination for retinoblastoma activity, and strong myopia prediction using fundus images. Several studies supported feasibility in neonatal, school, and community workflows using smartphone-based imaging, task-shifted operators, tele-referral, and human-in-the-loop review. However, external and temporal validation, calibration, patient-level reporting, subgroup and fairness assessment, and economic evaluation were limited. Conclusions: AI-enabled tools show promise for supporting selected paediatric ophthalmic screening, triage, and surveillance pathways, particularly when combined with image-quality control, explicit escalation, and human oversight. However, confidence in the reported performance is limited by single-centre studies and enriched samples, small numbers of clinically important cases, heterogeneous analytical units, potentially optimistic aggregation procedures, limited external or temporal validation, incomplete calibration, and absent fairness analyses. Routine autonomous implementation remains premature. Full article

Human pose estimation (HPE) has become a core technology for human-centered visual understanding and is widely used in sports analysis, rehabilitation assessment, human–computer interaction, autonomous driving, and industrial safety monitoring. Although benchmark-driven research has substantially improved pose estimation accuracy, strong performance on standard datasets does not necessarily ensure reliable deployment in real-world environments. Practical HPE systems are commonly deployed in camera-based, mobile, wearable, and edge-computing environments. In these settings, sensor quality, device placement, frame rate, illumination conditions, hardware limitations, temporal instability, and application-specific reliability requirements jointly influence system performance. Using a structured narrative synthesis of representative literature, this review examines HPE from a deployment-oriented perspective instead of treating pose estimation solely as a model-level task. The review first summarizes the major benchmark-to-deployment gaps, including data, system, task, and temporal gaps; it then analyzes the end-to-end HPE pipeline, including image or video acquisition, human detection, region localization, pose inference, tracking, temporal smoothing, coordinate decoding, and task-level decision-making. Representative CNN-based, Transformer-based, temporal modeling, and lightweight optimization methods are reviewed in terms of deployment cost, optimization potential, and application suitability. This review also discusses practical application requirements and future research directions related to deployment-aware evaluation, system-level optimization, efficient temporal modeling, and application-level reliability. By linking benchmark performance with practical sensing and system constraints, this review provides a structured reference for the selection, evaluation, and deployment of HPE methods in real-world environments. Full article

Global Navigation Satellite Systems (GNSS) are incompatible with the underwater domain as radiofrequency signals cannot penetrate the water column, leaving Autonomous Underwater Vehicles (AUVs) reliant on dead-reckoning systems that accumulate positional errors over time. When AUVs surface to reset their navigation, they face another challenge: GNSS itself is increasingly vulnerable to jamming and spoofing in contested environments. Automated Celestial Navigation (CN) has emerged as a promising alternative to other navigation methods, making it possible to derive the absolute position from observations of celestial bodies, entirely independent of human-made signals. This work provides a state-of-the-art review of automated CN technologies, focusing on the literature from 2020 onwards. The review covers Solar Tracking Sensors (STSs), star trackers and horizon detection algorithms and assesses their suitability for AUV integration through a structured SWOT analysis. Following this, a conceptual CN system based on Sunto’s STS is developed for the Light Autonomous Underwater Vehicle platform employing a proposed ten-step integration methodology. Computer-Aided Design models illustrate the conceptual setup, though hydrodynamic/structural verification remains subject to future work. Results suggest that solar-based CN can serve as a periodic absolute position corrector within a hybrid AUV navigation architecture, without requiring satellite infrastructure, which contributes towards a more resilient AUV navigation. Full article

Unprovoked shark bites are one of the most recognised human–wildlife conflicts and present a significant concern for beach safety. Lethal methods of shark mitigation have previously been implemented to reduce the risk of such incidents; however, due to their destructive impacts on vulnerable marine wildlife, non-lethal approaches are increasingly preferred. In recent years, drones have emerged as an effective, minimally invasive tool for real-time shark surveillance in surf zones. Drones are also used to collect valuable data on shark ecology and behaviour in nearshore environments, which can inform evidence-based policies. This review examines the utility of drones for shark surveillance programs by identifying key operational parameters and associated challenges of drone-based methods. We investigate emerging technologies, including long-range drones, remotely operated or autonomous flight missions, and the use of artificial intelligence for shark detection and species identification. We also outline current drone licensing, laws, and regulations, noting that these vary across administrative regions (i.e., countries and states). Overall, this review provides insight into the expansion of drone-based shark surveillance in nearshore areas and its potential to enhance beach safety, support management decisions, and advance scientific knowledge without negatively impacting shark populations. Full article

Automating activities in unstructured environments, such as construction sites, has been challenging due to unpredictable events, limiting robot adoption compared to structured settings. Recently, pre-trained foundation models, particularly Large Language Models (LLMs), have shown promise in addressing this challenge through superior generalization capabilities. This study employed a multi-dimensional method that systematically reviews the field from different perspectives of foundation models in robotics and unstructured environments, and synthesizes them with deliberative acting theory. The findings revealed that LLMs’ linguistic capabilities are primarily used to improve perception and human–robot interactions in robotic tasks, while applications in project management, safety, and natural hazard detection are the most utilized applications of foundation models in unstructured environments. Our synthesis shows an empirical gap in the field where fewer identified studies within unstructured environments validated their foundation model applications using physically deployed robots. We positioned the current state-of-the-art on a five-level automation scale of conditional automation. These findings inform future scenarios, challenges, and solutions toward autonomous safe unstructured environments. Our study serves as a benchmark to track our progress toward that future. Full article

Simultaneous Localization and Mapping (SLAM) is essential for autonomous systems navigating in human-centric environments, yet conventional systems fail when people and objects move through the scene. This paper introduces SAR-SLAM (Semantic-Aware Recognition SLAM), an RGB-D SLAM framework that robustly handles dynamic scenes containing moving people and objects using dual semantic geometric processing. First, we employ YOLOv8-based semantic segmentation to identify dynamic objects and generate initial detection masks. Second, we apply RANSAC-based Homography analysis to perform geometric motion verification, distinguishing truly moving objects from stationary ones by analyzing feature correspondence patterns. Third, an adaptive fusion mechanism combines both semantic and geometric evidence while incorporating temporal consistency and coverage constraints to maintain system stability. The system is implemented as a modular ROS2 package, enabling smooth integration with robotic systems and compatibility with existing navigation frameworks. SAR-SLAM reduces Absolute Trajectory Error by up to 96% over ORB-SLAM3 on the dynamic sequences of the TUM RGB-D benchmark, and remains competitive with state-of-the-art dynamic SLAM methods across a range of dynamic scenarios. Full article

This paper presents a systematic literature review examining how blockchain technologies can enhance the security and performance of multi-cloud systems. Multi-cloud architectures offer resilience, scalability, and flexibility; however, they also pose complex security challenges related to APIs, service-level agreements (SLAs), orchestration, and authentication. The promise of blockchain technology to improve the security and transparency of numerous applications, including cloud storage systems, has attracted considerable attention in recent years. Much research has focused on decentralized storage in cloud environments, spanning supply chains, FinTech, healthcare, and education. Still, the integration of blockchain with the cloud and its potential to enhance security and performance warrant an in-depth study. Using the PRISMA methodology, a structured search was conducted across six major scientific databases, including IEEE, ACM Digital Library, ScienceDirect, Scopus, Web of Science, and IJIMAI. Twenty-four primary papers published between 2019 and 2025 were selected for analysis after clear inclusion and exclusion criteria were applied. This review examines the security dimensions in multi-cloud environments—architectural vulnerabilities, API security, authentication, orchestration and automation vulnerabilities, SLAs, and cybersecurity compliance issues—in relation to blockchain technology. Based on the identified gaps, we propose a SOC-LLM-augmented security architecture that integrates blockchain-based evidence integrity, statistical anomaly detection, machine learning, large language models, and autonomous AI agents to enable intelligent attack detection and response. The proposed framework introduces specialized agents for detection, correlation, threat intelligence retrieval, blockchain evidence validation, explanation generation, and response planning. The analysis shows that integrating SOC-LLM capabilities with blockchain can move multi-cloud security from passive auditability toward proactive, explainable, and human-in-the-loop cyber defense. Finally, this paper discusses open challenges, including LLM hallucination, data scarcity, real-time scalability, evaluation standardization, and trustworthy deployment in critical multi-cloud infrastructures. The study’s conclusion highlights research gaps and suggests future lines of inquiry concerning scalable blockchain architectures and the incorporation of AI for proactive cloud security monitoring. Full article

Autonomous vehicle (AV) technologies have advanced rapidly in recent years, driving an increasing demand for large-scale, high-quality annotated data. However, collecting and annotating real-world pedestrian video datasets is time-consuming, costly, and often insufficient to cover rare and safety-critical scenarios. Recent world foundation models have demonstrated impressive capabilities in generating realistic videos, yet their suitability for safety-critical autonomous driving applications remains largely unexplored. In this work, we investigate whether current world foundation models can generate driving scenarios that are sufficiently realistic and behaviourally consistent for autonomous driving research. We conduct a case study centred on pedestrian–vehicle interactions captured from ego-vehicle dashcam viewpoints, where subtle behavioural and geometric errors can have significant safety implications. To support this investigation, we develop SynPeDAS, an open research framework comprising a collection of synthetic pedestrian-interaction videos, a reusable generation pipeline for transforming real-world driving footage into synthetic scenarios, an automated evaluation suite, and downstream demonstration code. Through quantitative evaluation and structured human assessment, we identify several recurring failure modes, including dynamic misalignment, depth drift, and object persistence inconsistencies. More importantly, we find that commonly used evaluation metrics frequently exhibit ceiling effects and weak alignment with human judgement, limiting their ability to detect safety-critical behavioural errors. These findings indicate that, despite high perceptual realism at the frame level, current generative world models and existing evaluation methodologies remain insufficient for capturing physically grounded motion and task-critical semantics. Consequently, significant challenges remain before world model-generated videos can be considered reliable for safety-critical autonomous driving applications. SynPeDAS provides an open platform for systematically studying these challenges and developing improved generation and evaluation methods. Full article

Educational environments, particularly those with limited resources, require affordable mobile robots capable of combining human–robot interaction, autonomous assistance, and academic support without continuous dependence on cloud services. This work presents a low-cost ROS2-based mobile robot implemented on a Raspberry Pi 4B to provide educational assistance in Spanish within controlled classroom environments. The system integrates voice interaction, text-to-speech synthesis, YOLOv8n-based object perception, a specialized door detection model, ultrasonic and inertial sensing, differential-drive control, and a hybrid natural language processing architecture based on semantic caching, local inference, and optional cloud connectivity. Two task-dependent operating modes, education and navigation, selectively activate ROS2 nodes to reduce computational load and energy consumption. Experimental tests conducted in a university classroom evaluated speech recognition, vision models, natural language processing alternatives, sensor behavior, and battery life. The speech recognition module achieved 98% accuracy under both quiet and noisy conditions. YOLOv8n achieved an F1-score of 0.975 for common classroom objects, while the specialized door detector achieved 100% recall with 58.7% precision. The semantic cache correctly resolved recurrent academic queries in the exact-match evaluation, with an average latency of 3.8 s, reducing the need for external language models in known-question scenarios. The robot operated for 96 min in education mode and 75.6 min in navigation mode. These results demonstrate that Spanish voice interaction, reactive navigation, academic question answering, and resource-aware operation can be integrated into a single low-cost edge robotic platform for educational environments. Full article

Visible and infrared (IR) image fusion has become an important strategy for improving computer vision performance under low illumination, occlusion, and some poor-visibility conditions. By integrating complementary textural information from visible images with thermal or IR cues, VIR fusion can enhance object localization, detection robustness, and facial-expression recognition (FER). This review examines VIR fusion techniques and datasets for computer vision applications, with object detection (OD) considered as a relatively mature scene-level task and FER considered as an emerging human-centered application. It summarizes major multimodal datasets, compares early-fusion approaches, including sensor- and feature-level fusion, with late-fusion approaches, including score- and decision-level fusion, and discusses representative machine learning and deep learning methods. The review also evaluates commonly used performance metrics and identifies current limitations, including dataset imbalance, sensor misalignment, limited demographic diversity in facial-expression datasets, computational complexity, and weak real-time generalization. Finally, key application areas, including surveillance, healthcare, remote sensing, autonomous systems, and human–computer interaction, are discussed. This review highlights the need for better-aligned multimodal datasets, standardized evaluation protocols, lightweight fusion architectures, and robust models capable of operating in dynamic real-world environments. Full article

Crime surveillance in urban environments faces increasing challenges due to dynamic conditions and the demand for real-time monitoring. This paper investigates the use of video data from autonomous vehicles to enhance situational awareness in public spaces through deep learning models optimized for edge processing. High-resolution vehicle-mounted cameras serve as mobile surveillance units capable of real-time object detection, human action recognition, and anomaly detection, bridging the gap between autonomous mobility and urban monitoring. Building on this vision, we introduce ITS-Vision, a generic framework that operationalizes these use cases, enabling autonomous vehicles to function as mobile, context-aware sensing platforms. To validate this approach, we develop prototypes for key ITS-Vision components: a fight detection module using a fine-tuned X3D model, suspect identification via MediaPipe for detection combined with FaceNet for embedding extraction, and a dangerous items detection module using a fine-tuned YOLOv11n model. Due to the limited availability of real-world autonomous vehicle video datasets, experiments were conducted in controlled laboratory environments, demonstrating the feasibility of the proposed architecture and algorithms under simulated conditions. Future work will focus on collecting dedicated datasets and advancing the models toward deployment in real urban scenarios. Full article

Process industries, including chemical, petrochemical, energy, and wastewater treatment systems, operate under high-dimensional, nonlinear, dynamic, and safety-critical conditions. Data-driven process monitoring and fault detection and diagnosis (FDD) have progressed from multivariate statistical monitoring to machine learning and deep learning. Yet many deployed or prototype systems still behave mainly as fault classifiers: they detect deviations, but offer limited causal explanation, weak integration of plant knowledge, and insufficient support for safe operator action. Recent advances in large language models, retrieval-augmented generation, digital twins, explainable artificial intelligence, and multi-agent systems make it timely to revisit FDD as an agentic decision-support workflow. This focused review examines how agentic AI can support process-industry monitoring and diagnosis by integrating process data, engineering knowledge, model outputs, and safety constraints. We synthesize established FDD foundations, deep-learning-based FDD, process-safety context, bridging technologies, and emerging LLM- and agent-based studies. The review argues that the near-term value of industrial agents lies not in unrestricted autonomous plant control, but in safety-aware, explainable, and human-in-the-loop decision support. We propose a process-industry-oriented taxonomy of agents, summarize enabling technologies and representative application settings, and identify evaluation criteria, benchmark requirements, limitations, and deployment conditions for trustworthy industrial agents. Full article

Industrial quality control is a critical component of modern manufacturing, as defects can lead to significant economic losses and safety risks. Traditional inspection methods, largely reliant on human operators or rule-based systems, often suffer from inconsistency, limited scalability, and reduced accuracy in complex environments. Recent advances in artificial intelligence (AI), particularly in deep learning and computer vision, have enabled automated defect detection and classification with unprecedented performance. This paper provides a comprehensive review of AI-based image processing techniques for industrial quality control, covering classification, detection, and segmentation approaches. Key applications across manufacturing sectors are discussed, alongside current challenges such as data scarcity, real-time implementation, and model generalisation. Furthermore, this paper explores emerging trends toward autonomous inspection systems, integrating real-time analytics, edge computing, and intelligent decision making. The insights presented aim to guide future research toward robust, scalable, and fully automated quality control solutions in smart manufacturing environments. Full article

Saudi Arabia’s audit market faces three governance challenges that existing frameworks may not fully address. These challenges concern a potential regulatory gap around autonomous AI accountability, a trust dimension that standard technology-adoption models may not fully capture, and limited mechanisms for independently verified ESG assurance under Vision 2030. This study adopts a conceptual design approach within the design science research tradition and proposes the CMA Agentic AI Platform as a practical response to these challenges. The platform comprises two segments. Segment 1 deploys autonomous drone swarms to verify corporate assets across four audit tasks—asset valuation, ESG compliance, anomaly detection and construction progress—using deep learning, thermal imaging and social-media cross-referencing. Segment 2 continuously monitors discretionary accruals and uses objective earnings-management data to inform auditor assignment and rotation decisions. This approach replaces subjective reputational assessments with transparent, quantifiable governance criteria. The platform is governed through the Triadic Agentic Framework, which extends classical agency theory by distributing authority across the Principal, the Human Agent and the AI Agent. The framework also operationalises Trust Expectancy as the primary adoption condition. The evidence base draws on two complementary streams: a PRISMA-guided systematic review and bibliometric analysis of thirty-nine peer-reviewed studies, and a documentary analysis of four national agentic-AI regulatory frameworks (SDAIA, MDDI/IMDA, NIST and ICO). The study contributes the concept of Algorithmic Accountability as a distinct governance domain, the Triadic Agentic Framework as an operational architecture for autonomous regulatory monitoring, and a reframing of the UTAUT trust construct for agentic-AI adoption in mature professional contexts. The platform converts theoretical governance into a regulatory architecture with direct implications for concentrated capital market regulators. Full article