Search Results (90)

In horticulture, a delayed shipment or failed food-safety check can rapidly cut off market access, yet strategic risk management remains weakly developed. This paper examines how artificial intelligence (AI), agritech, and business model innovation can reduce uncertainty and strengthen resilience across horticultural value chains. Using a structured conceptual review, it integrates six literature streams: agricultural risk, supply-chain risk management, strategic and enterprise risk management, farming-system resilience, digital innovation and AI, and business model innovation. The synthesis connects risk exposure to five digital functions (sensing, predicting, verifying, coordinating, and deciding) to dynamic capabilities, business model and governance conditions, and resilience outcomes. It develops five connected propositions concerning the conversion of information into action, credible digital verification, adoption conditions, market-access resilience, and business-model-enabled shared capability. Market-access resilience is introduced as the capacity to maintain or restore verified quality, food-safety assurance, traceability, compliance, and buyer relationships during disruption. The propositions are theoretically derived and empirically informed, providing a foundation for further empirical validation, particularly of the complete market-access and business model relationships. The findings imply that digital technologies contribute to resilience only when supported by coordinated action, sound data governance, human oversight, and viable, inclusive delivery models. Full article

Digital-twin-enabled human–machine collaboration (HMC) has increasingly been proposed as a system-level approach for connecting human operators, robots, sensors, artificial intelligence modules, and manufacturing resources. However, the literature varies substantially in what is called a digital twin, how physical and virtual models are coupled, whether models are updated from physical data, and how far systems have progressed beyond simulation or controlled laboratory demonstrations. This structured integrative review examines the conditions under which a digital twin can function as an integration layer for HMC in sustainable smart manufacturing, rather than assuming that such integration is already established industrial practice. The literature corpus was assembled through searches of the Web of Science Core Collection, Scopus, and IEEE Xplore, complemented by Google Scholar-based citation tracking and backward and forward citation tracing. The core search focused on studies published from 1 January 2020 to 5 August 2026, while earlier seminal studies were retained to support definitions and historical context. Studies were screened using explicit criteria for manufacturing relevance, physical–virtual coupling, state synchronization or model updating, feedback capability, and validation setting, and were critically coded by model type, integration mechanism, deployment maturity, and sustainability evidence. The review compares multimodal perception and human-state modeling, intention understanding and augmented interaction, task allocation and shared planning, digital-twin architectures, adaptive control and safety verification, and human–AI decision-making. The evidence indicates that digital twins are promising as coordination and verification layers, but many reported systems remain conceptual, simulation-based, or limited to controlled physical prototypes. Key barriers include model fidelity, online model updating, real-time synchronization, cross-platform interoperability, safety assurance, human-data governance, and the limited availability of directly measured sustainability outcomes. Future work should prioritize validated hybrid models, traceable model-update mechanisms, staged virtual-to-physical deployment, interoperable data contracts, and longitudinal evaluation of technical, human, economic, and environmental performance. Full article

Distributed-drive electric vehicles (DDEVs) enable independent wheel-torque control, providing flexibility to improve energy efficiency and vehicle stability. However, tire–road adhesion, motor and battery capabilities, and actuator availability constrain these objectives, which may conflict under low-adhesion conditions, high-power acceleration, emergency braking, and combined longitudinal–lateral maneuvers. This paper provides a structured review of coordinated torque-allocation strategies for balancing energy efficiency and stability in DDEVs. Existing research is examined in terms of regenerative braking, tire-slip energy-loss reduction, and stability control under longitudinal, yaw, and combined conditions. Control approaches are classified as rule-based, stability-region-based, mode-switching, multi-objective optimization and predictive control, state-adaptive dynamic-priority coordination, and learning-based safety-hybrid methods. These approaches differ in real-time performance, constraint handling, adaptability, interpretability, and engineering maturity. A hierarchical hybrid architecture integrating rule-based supervision, state assessment, constraint-aware optimization, and learning-based enhancement appears more suitable for practical deployment than a single algorithm or fixed-weighting scheme. Key challenges include dynamic stability-boundary estimation, safety-assured coordination, multi-actuator fault tolerance, real-time implementation, and standardized vehicle-level validation. This review provides guidance for coordinated control-system development and future research on DDEVs. Full article

Hazard analysis and critical control point (HACCP) systems provide a preventive framework for food safety by implementing quality assurance plans, continuous monitoring, corrective actions, and risk mitigation strategies at critical control points throughout food processing, including dairy products such as ice cream. Artificial intelligence (AI) is increasingly transforming food safety management by enabling real-time monitoring, predictive analytics, and automated decision-making within food processing systems. This review critically examines the integration of AI technologies into HACCP systems for ice cream manufacturing, with an emphasis on improving hazard detection, process control, traceability, and the efficiency of corrective actions. The review evaluates the application of Internet of Things sensors, computer vision, and machine learning-based predictive monitoring systems across critical processing stages, including raw material reception, pasteurization, continuous freezing, and hardening/storage. Compared to conventional HACCP systems, AI-assisted technologies offer greater capabilities for anomaly detection, predictive maintenance, automated verification, and data-driven risk management. Nevertheless, their industrial implementation remains constrained by data quality limitations, infrastructure cost, cybersecurity risks, regulatory uncertainty, and limited model explainability. Accordingly, this review highlights key research gaps related to industrial scalability, validation under dynamic processing conditions, and the scarcity of ice cream-specific AI datasets. Finally, the review identifies future research directions and emerging opportunities for applying AI technologies in food processing and quality control systems, providing a framework for the evolution of intelligent HACCP systems in frozen dairy manufacturing. Full article

With the vigorous development of offshore energy and mining, offshore fracturing and deep-sea mining necessitate the ship-to-ship and platform-to-ship transfer of solid particles via floating hoses. However, traditional floating hoses designed for oil transportation are inadequate for the long-term and stable conveyance of granular materials, as solid particles are prone to deposition and blockage under excessive bending. Additionally, in large-scale offshore fracturing operations, tension fluctuations in high-pressure hoses accelerate hose wear and compromise structural integrity. To address these challenges, this study proposes a systematic framework integrating neural network prediction with a feedforward–feedback composite control strategy. Spatial Attention–Convolutional Neural Network (SA-CNN) achieves the highest prediction accuracy and the strongest generalization capability across all operating conditions. Then, a feedforward–feedback composite control strategy is formulated, where the feedforward component is derived from SA-CNN predictions and the feedback component is provided by a PID controller, with an adaptive weighting mechanism adjusting their contributions based on prediction confidence. A curvature safety constraint is also incorporated to prevent excessive bending. The results show that the composite control strategy achieves the highest peak tension reduction, while achieving the lowest RMSE. Unlike pure PID, which introduces severe oscillations, the composite control strategy converges smoothly, confirming that feedforward prediction effectively suppresses feedback-induced oscillations. This study provides a theoretical foundation and a practical solution for the safety assurance of floating hoses in high-pressure fracturing fluid delivery applications and offshore solid particle transshipment. Full article

The shift from passive predictive models to autonomous agents capable of tool use and multi-step planning moves the AI safety landscape from prediction error to control failure: small misjudgements become irreversible actions, and risks compound across long horizons and populations of interacting systems. We present a structured review and taxonomy of open scientific problems in agentic AI safety, mapped explicitly onto the EU AI Act and the NIST AI Risk Management Framework. The corpus follows a PRISMA-ScR scoping review, assembled through anchor-based citation chaining and curated reading lists across arXiv, the major machine-learning conferences, and selected security and fairness venues, with a primary March 2026 search cut-off (extended to May 2026 during revision for a small number of high-relevance governance and agentic-safety sources), explicit eligibility criteria, and an analytical distinction between open scientific problems and deployment risks. The taxonomy identifies eight problem families spanning reinforcement-learning policies and language-model planners: goal specification, inner alignment, safe learning and robustness, scalable oversight, interpretability, tool-use security, multi-agent safety, and evaluation and assurance. Mapping these onto the two frameworks shows close alignment for some families and notable absences for others, with multi-agent safety surfacing as a regulatory gap. We add a per-family research roadmap with concrete milestones and a practitioner-facing deployment-posture triage, arguing that progress on inner alignment, interpretability for deceptive-alignment detection, and multi-agent safety would most directly reduce compliance uncertainty. Full article

Additive manufacturing (AM) has expanded the design space of biomaterials for biomedical engineering, enabling patient-specific geometries, controlled porosity, multi-material constructs and cell-compatible fabrication. However, clinical translation remains constrained by a mismatch between fabrication capability and biological, mechanical and regulatory performance. Unlike reviews focused on individual material families, isolated AM routes or specific applications, this review interprets AM of biomaterials as an integrated biomaterial–process–structure–property–translation ecosystem. It examines how material chemistry, feedstock state, printing route, architecture, post-processing and biological response jointly determine the reliability of acellular and cellular constructs, with particular emphasis on clinically relevant performance, reproducibility and long-term safety. Metallic, ceramic, polymeric, hydrogel-based and composite biomaterials are analysed alongside binder jetting (BJ), directed energy deposition (DED), material extrusion (ME) and jetting (MJ), powder bed fusion (PBF), VAT photopolymerisation and bioprinting. The review identifies recurring challenges across routes, including restricted material–process compatibility, limited prediction of process–structure–property relationships, post-processing-induced changes in biological performance, insufficient standardisation of printability and biofunctionality metrics, incomplete validation of cell-laden and vascularised constructs and weak transfer of laboratory protocols to clinically robust workflows. The main conclusion is that progress will depend less on expanding printable geometries alone and more on integrated optimisation of materials, processing windows, structural fidelity, biological validation, quality assurance and translational readiness. This ecosystem-level perspective provides a framework for evaluating limitations and defining future priorities in AM-based biomaterials for regenerative medicine, implants and precision biomedical engineering. Full article

Wide-bandgap (WBG) power converters impose stringent requirements, including high-frequency switching, strongly nonlinear dynamics, and limited computational time, which constrain conventional control and artificial intelligence (AI)-based approaches and hinder their practical deployment. Existing review studies have primarily focused on algorithmic structures or performance, while systematic analyses from a deployment feasibility perspective under hardware constraints remain limited. This paper examines AI applications in WBG power converters from a system-level deployment perspective and analyzes existing studies based on implementation feasibility. After outlining the physical characteristics and control requirements of WBG devices, it reviews AI-based modeling, AI-assisted model predictive control (MPC), and reinforcement learning (RL)-based direct control. These approaches are evaluated in terms of computational complexity, real-time feasibility, out-of-distribution (OOD) generalization, and integration with conventional control frameworks. Key deployment challenges, including safety-constrained RL, sim-to-real transfer, and field-programmable gate array (FPGA)/embedded implementation, are treated as core analytical dimensions. To support this assessment, this review introduces an AI Deployment Readiness framework organized around four analytical dimensions: (1) modeling accuracy, (2) safety assurance, (3) sim-to-real transfer capability, and (4) hardware implementability. Using this framework, prior studies are reassessed, and its applicability is further discussed for applications such as fault diagnosis and remaining useful life (RUL) prediction. The analysis identifies key bottlenecks and clarifies deployment-relevant considerations for high-frequency WBG systems. Full article

Rising atmospheric CO₂ levels pose a critical challenge to achieving global sustainability targets. Geological carbon sequestration (GCS) offers a long-term solution for reducing greenhouse gas emissions, but its large-scale deployment faces limitations in cost, uncertainty, and operational risk. Recent advances in machine learning (ML) present transformative opportunities to enhance every stage of the carbon capture and storage (CCS) lifecycle, from pre-injection evaluation to post-injection monitoring. This review systematically examines ML integration in CCS applications, emphasizing roles in geological characterization, injection optimization, plume prediction, and leakage detection. It provides a structured overview of ML methodologies including Random Forest, Support Vector Regression, and XGBoost, along with emerging deep learning models used for anomaly detection and uncertainty quantification. Experimental insights, monitoring techniques, and real-time data applications are summarized to illustrate ML’s capability in accelerating simulations, reducing costs, and increasing safety assurance. Furthermore, real-world case studies such as Sleipner (Norway), Illinois Basin–Decatur (USA), Boundary Dam (Canada), Gorgon (Australia), and Quest (Canada) demonstrate how ML has enhanced performance, predictive accuracy, and storage reliability in field-scale CCS projects. The review concludes by identifying existing challenges, data scarcity, interpretability, and regulatory integration, and proposes a unified ML framework for scalable, autonomous, and secure CO₂ storage. Overall, this study provides a comprehensive roadmap for leveraging artificial intelligence to achieve reliable, cost-effective, and sustainable carbon management solutions aligned with global net-zero objectives. Full article

To address the demand for high-precision, high-efficiency, and standardized hydrographic information in intelligent shipping, this study systematically investigates key technologies for high-resolution hydrographic data parsing and intelligent information services. Focusing on the East China Sea, a space–air–ground integrated monitoring data access system is established. A hybrid data assimilation method combining four-dimensional variational (4D-Var) and ensemble Kalman filter is adopted to realize quality control, deep fusion, and optimal state estimation of multi-source heterogeneous hydrographic observations. A hybrid tidal harmonic response model is further developed to improve the refined forecasting accuracy of tide levels and ocean currents. A hierarchically decoupled system architecture is designed, and modules for data production, sharing, exchange, and visualization are developed in compliance with the international S-100 standard. By integrating hybrid spatiotemporal indexing, multi-level caching, and intelligent query optimization, the system achieves low-latency and high-concurrency service capabilities. Experimental results show that, compared with conventional models, the proposed framework reduces tidal forecast RMSE by approximately 15.8% under extreme weather, raises the continuity index of current vectors to 0.93, and cuts the S-100 product generation latency to less than 30 s. This research establishes a full-chain technical system from data parsing and product generation to intelligent services, providing a reliable technical support platform for ship intelligent navigation, dynamic route planning, and maritime safety assurance. Full article

Traditional coal mining methods have led to prominent issues of coal resource waste and large-scale solid waste emissions. The integrated “excavation–backfilling–retention” mining technology for inter-section coal pillars and gate roads is one of the key technologies to solve these problems. However, the excavation and mining process associated with this technology imposes higher requirements on the ventilation system. Aiming at addressing the ventilation challenges existing during the implementation of the “excavation–backfilling–retention” method, research on ventilation safety assurance technology for inter-section coal pillars was carried out. Using COMSOL5.5 software, a full-stage ventilation system design model was constructed, adopting a ventilation mode that combines full-air-pressure ventilation with auxiliary local ventilation. The dynamic variation characteristics of the ventilation system under the “excavation–backfilling–retention” method and its capability to prevent and control the risks of O₂ and CO gas accumulation and coal spontaneous combustion were studied. The results show that during the bypass excavation period, the air supply from the auxiliary fan is sufficient, and during the excavation period for the two gate roads, due to the increased ventilation distance, insufficient airflow occurs near the heading face, accompanied by temperature rise, O₂ concentration decrease, and local CO accumulation, posing risks of coal spontaneous combustion and toxic gas accumulation. During the inter-section coal pillar excavation period and the cyclic operation period, after the full-air-pressure ventilation system is established, the airflow becomes stable, ventilation resistance decreases, and both temperature and gas concentrations are controlled within safe limits. However, in the corner areas, auxiliary local ventilation measures are still required due to insufficient O₂ and CO accumulation. The study verifies the feasibility and safety of the integrated “excavation–backfilling–retention” ventilation system, providing a safe ventilation approach for the integrated mining method and supporting the green mining of coal mines and the synergistic development of coal-based solid waste resource utilization. Full article

Industrial automation systems are increasingly cyber-physical, interconnected, and software-dependent, which expands both their operational capability and their cybersecurity exposure. This article reports a systematic literature review, conducted following the PRISMA 2020 guidelines, of cybersecurity requirements and certification standards in industrial automation, with emphasis on Industrial Control Systems (ICS), Supervisory Control and Data Acquisition (SCADA), Programmable Logic Controllers (PLCs), and Industry 4.0 contexts. From 3570 records identified across five academic databases, 75 studies were retained after duplicate removal, title and abstract screening, and full-text eligibility assessment. The included studies were analyzed along three dimensions: cybersecurity requirements, standards and certification, and application context. Quantitative synthesis shows that network segmentation, intrusion detection, secure communication, access control, lifecycle security, and safety–security coordination are the six most frequently emphasized requirement categories, and that ISA/IEC 62443, ISO/IEC 27001, NIST SP 800-82, and NERC-CIP are the four dominant certification frameworks. The review identifies four critical gaps between technical cybersecurity requirements and certification practice and proposes an integrated mapping framework linking requirement categories, standards, and application contexts. The findings indicate that effective industrial cybersecurity assurance depends on a layered compliance architecture rather than on dependence on any single framework. Full article

The rapid evolution of Large Language Models (LLMs) has exposed limitations of static, accuracy-oriented benchmarks and increased the need for evaluation frameworks that distinguish among capabilities and benchmark quality. This survey analyzes 63 LLM benchmarks spanning 2012–2026 and organizes them into a taxonomy of six capability dimensions and 20 operational subcategories. We also propose the Benchmark Quality Assurance Index (BQAI), an AHP-weighted composite framework for assessing the scientific quality of benchmarks across seven dimensions related to annotation, clarity, standardization, reproducibility, robustness, coverage, and fairness. The BQAI is applied to 30 representative benchmarks, corresponding to 48% of the 63-benchmark corpus, with three-evaluator blinded scoring, formal inter-rater reliability validation $\left(\mathrm{ICC}(2,k)\right.$ and quadratic-weighted Cohen’s $\left.\kappa \right)$, and Monte Carlo sensitivity analysis $\left(n=1000\mathrm{trials},\pm 10\%\mathrm{to}\pm 50\%\mathrm{weight}\mathrm{perturbation}\right)$. In addition, we synthesize public performance results for 16 models across 10 benchmarks to examine saturation trends and reporting gaps. The analysis indicates that benchmark usefulness varies substantially across evaluation settings, that several established benchmarks are becoming less discriminative for frontier models, and that important gaps remain in safety, agentic, and cross-cultural assessment. Together, the taxonomy, BQAI, and saturation analysis provide a structured perspective on the current LLM benchmark landscape and on priorities for more rigorous evaluation. Full article

Anthropic’s April 2026 Claude Mythos Preview release established a new operational threat category: frontier AI systems whose extended-context reasoning, recursive self-correction, native system-tool integration, and agentic scaffolding render dominant AI safety paradigms—RLHF, output filtering, contractual access vetting, human-in-the-loop supervision—insufficient as sole controls. This paper develops a defense-in-depth reference architecture against that category, structured around four named contributions: a five-indicator operational definition of the Mythos-class (capability conjoined with scaffold, access pattern, autonomy depth, and persistence); the Mythos-Class Posture Rubric (MCPR), a three-tier detection framework spanning evaluation, deployment, and runtime with explicit routing to mitigation layers; a four-layer mitigation stack comprising the Vetted-Access Operational Pattern (VAOP), Authority-Bound Output Release (ABOR) cryptographically grounded in FIPS 203/204/205 post-quantum primitives, and the Compute-Plane Isolation Profile (CPIP); and an integrated architecture that crosswalks to the NIST AI Risk Management Framework, NIST Cybersecurity Framework 2.0, and CISA Zero Trust Maturity Model 2.0. The architecture is applied to three deployment surfaces—post-quantum cryptography migration, federal AI supply-chain assurance, and critical-infrastructure operational technology defense—demonstrating that the four contributions generalize across heterogeneous operational contexts. The contribution is a reference design rather than a deployed system; limitations, falsifiability criteria, and a research agenda for empirical refinement are developed. Full article

Background: Smart offloading technologies enable the real-time, objective monitoring of adherence in patients with diabetic foot ulcers (DFUs). Although remote tracking may reinforce adherence and improve wound healing, effectiveness depends on sustained device use, particularly as devices are often removed during rest periods. Real-time, behavior-contingent feedback informed by sensor data, including AI-supported messaging capable of detecting nonadherence, may enhance reinforcement. However, the feasibility and behavioral impact of such strategies remain unclear. Methods: We conducted a prospective feasibility case series nested within a larger DFU cohort of 210 participants, enrolling eight adults with active DFUs. Participants used a sensor-integrated offloading device paired with a smartwatch (SmartBoot) and a mobile application (CORA) that delivered notifications to their smartphones. Notifications were either schedule-based or context-aware, using real-time SmartBoot data to generate personalized messages. The primary outcome was a sensor-detected transition from nonadherent to adherent offloading within 60 min. Results: A total of 130 notifications were delivered, with 125 included in the behavioral response analysis. Context-aware notifications demonstrated higher transition rates than schedule-based notifications. Adaptive Reinforcement yielded the highest response rate (77.4%, 24/31), followed by Clinical Course Correction (71.4%, 20/28), whereas Safety and Technical Assurance (40.7%, 11/27) and Motivational Coaching (30.8%, 12/39) showed lower response rates. Conclusions: Real-time, context-aware feedback is feasible and associated with improved short-term adherence, supporting evaluation in larger trials. Full article