Search Results (5,614)

Ground settlement induced by shield tunnel boring machine (TBM) excavation is a major geotechnical concern in urban tunneling because it may affect the safety of adjacent structures and underground infrastructure. In this study, machine learning models were developed to predict the maximum settlement induced by shield TBM excavation using a three-dimensional numerical analysis database comprising 320 simulation cases generated from combinations of tunnel diameter (D), ground elastic modulus (E), face pressure (FP), and backfill pressure (BP). Random forest (RF) and extreme gradient boosting (XGBoost) models were developed and compared with an existing regression-based settlement prediction equation. Predictive performance and generalization capability were evaluated using random split and GroupKFold validation techniques. Under random split validation, RF achieved the highest predictive performance, with a coefficient of determination of 0.997 and a root mean square error of 0.438 mm, followed by XGBoost. Both machine learning models outperformed the existing settlement prediction equation. However, model performance decreased substantially under GroupKFold validation, indicating limited generalization capability under unseen D–E grouped conditions. The results demonstrate that the developed machine learning models provide accurate predictions within the range of tunnel–ground conditions represented by the adopted numerical analysis database. The findings highlight the importance of evaluating both predictive performance and generalization capability, particularly when machine learning models developed from numerical analysis databases are applied beyond the conditions represented in the training database. Full article

Autonomous mobile robots are playing an increasingly significant role in modern smart orchards by supporting precision agricultural operations such as target-oriented spraying and autonomous harvesting. Nevertheless, achieving high-precision trajectory tracking and stable motion in complex, unstructured orchard environments remains challenging, because tracking deviations induced by uneven terrain and low-traction soil can directly affect operational safety and efficiency. To address this challenge, the present study proposes an adaptive tracking controller which integrates model-driven and data-driven approaches. Firstly, a six-state planar dynamic model based on Newton–Euler equations is established to describe motion characteristics. Secondly, an improved Particle Swarm Optimization (PSO) algorithm is employed for offline parameter optimization under representative operating conditions. The process thus engenders a mapping dataset that relates the real-time motion states of the orchard mobile robot to the optimized horizon parameters and weights. Finally, an Adaptive Neuro-Fuzzy Inference System (ANFIS) is trained using this dataset, enabling adaptive adjustment of MPC parameters according to the robot motion state. Simulation and experimental results demonstrate that, in Double-Lane-Change (DLC) and serpentine simulations, the proposed controller reduced lateral and heading Root-Mean-Square (RMS) errors to 0.0109 m/0.0081 rad and 0.0102 m/0.0117 rad, achieving reductions of 49.30–85.58% and 68.60–88.02% compared with Pure Pursuit, Stanley, Linear Quadratic Regulator (LQR), and traditional MPC, respectively. In orchard field tests with circular and Figure-8 trajectories at 0.3–0.6 m/s, the lateral RMS errors were recorded as 0.0112–0.0182 m and 0.0156–0.0262 m, respectively, corresponding to reductions of 46.94–61.52% relative to traditional MPC, while the heading RMS error remained below 0.0510 rad. These findings substantiate the efficacy of the proposed controller in enhancing the accuracy and adaptability of the system, thereby providing a resilient and precise control framework for operation within orchard environments. Full article

Accurate prediction of battery thermal runaway (TR) is a critical requirement for electric vehicle (EV) battery management systems (BMSs), as TR remains one of the most severe failure modes in lithium-ion batteries. Conventional neural network training methods may suffer from local optimum entrapment, slow convergence, and unstable performance when applied to nonlinear battery safety data. To address these limitations, this paper proposes a Hybrid Particle Whale Optimization Algorithm-optimized feedforward neural network (HPWOA-FNN) for continuous TR probability prediction and binary high-risk event classification using multivariate EV charging sensor data. The proposed HPWOA combines the rapid convergence capability of Particle Swarm Optimization (PSO) during the initial exploration phase with the exploitation and refinement capability of the Whale Optimization Algorithm (WOA) during the second phase. A global-best transfer mechanism is introduced at the PSO-WOA phase boundary to preserve the best solution identified during exploration and initialize the WOA leader, thereby improving convergence continuity and reducing premature stagnation. The model is evaluated using a 500-sample EV battery-charging dataset containing 12 electrothermal, electrical, mechanical, and environmental features. The proposed HPWOA-FNN outperforms standalone PSO-, WOA-, and Stochastic Fractal Search Algorithm (SFSA)-optimized FNN models across all regression metrics, achieving MSE = 0.000989, RMSE = 0.031442, MAE = 0.027250, R² = 0.9702, and MAPE = 3.8075%. For binary high-risk event detection, HPWOA-FNN achieves the highest AUC of 0.9817 and the lowest false-negative count, reducing missed high-risk events to 7 compared with 9 for PSO, 12 for WOA, and 17 for SFSA. Feature-importance analysis identifies maximum temperature and internal resistance as the dominant predictors, consistent with established thermal runaway mechanisms. The results demonstrate that HPWOA-FNN provides an accurate, interpretable, and computationally practical framework for EV battery thermal runaway prediction and BMS decision support. Full article

Galloping of overhead transmission lines threatens grid safety and requires non-contact measurement methods that can quantify three-dimensional (3D) motion from field video. This paper proposes a deep-vision and binocular-collaboration framework for dynamic conductor displacement measurement. The framework combines three components that are matched to the physical structure of transmission lines: adaptive image enhancement using Retinex illumination decomposition and Wiener blind deconvolution; a structure-prior dual-branch extraction module that uses an improved YOLOv11 keypoint branch for spacer-equipped sections and an improved U-Net branch with Dynamic Snake Convolution (DSC) and Strip Pooling for bare conductors; and stereo reconstruction with Kalman-filter-based temporal association for continuous trajectory estimation. Compared with the original submission, the revised manuscript further clarifies the real-video data acquisition, annotation procedure, camera synchronization, calibration workflow, training/testing independence, and runtime measurement protocol. Additional validation on a public real power-line image dataset is also reported. The proposed method achieves a Z-axis Root Mean Square Error (RMSE) of 24.5 mm for spacer sections in the controlled binocular field test, a dominant-frequency relative error below 3.5%, and 32 FPS on edge hardware when preprocessing, visual extraction, stereo projection, and temporal filtering are included. On the supplementary public power-line dataset, the segmentation branch obtains a Dice coefficient of 0.9039 and an IoU of 0.8395. These results indicate that the proposed framework reduces the depth-scale limitation of monocular vision and provides a practical quantitative tool for field galloping monitoring. Full article

As a core load-bearing component for railway vehicles, rails are largely responsible for the safety and stability of train operation, and their service performance is inherently governed by material microstructure. In this study, rails with varied pearlite interlamellar spacing were prepared and laid on field tracks for 8 months of service testing to investigate the influence of pearlite interlamellar spacing on rail wear and rolling contact fatigue (RCF). The results indicate that decreasing pearlite interlamellar spacing facilitated tread work hardening and reduced cumulative wear loss of rails. At the early service stage, rails with coarse pearlite lamellae exhibited earlier RCF crack initiation and longer crack morphologies, while rails featuring finer pearlite lamellae exhibited the latest-occurring crack initiation. With prolonged service duration, wear loss rose continuously, and the tread hardening rate first increased sharply and then tended to gradually become stable. Obvious differences in damage evolution were observed for rails with different pearlite interlamellar spacing. Coarse-lamellar rail suffered sparse short cracks dominated by wear; fine-lamellar rail developed dense fast-growing cracks controlled by RCF; and medium-lamellar rail achieved a relatively good balance between wear and RCF. A competitive relationship exists between wear and RCF during rail service. Reasonable regulation of pearlite interlamellar spacing facilitates a balanced evolution of wear and RCF, which provides a feasible microstructural optimization strategy for improving the service performance and service life of pearlitic rails. Full article

Hydrogen is increasingly recognised as one of the leading pathways for decarbonising the maritime sector. Proton exchange membrane fuel cell (PEMFC) hybrid propulsion is emerging as a promising low-emission technology; however, its safe deployment depends on marine engineers being trained to interpret and manage coupled hydrogen, fuel cell, battery, and electric propulsion systems. However, a critical training gap remains. Alternative fuel guidance identifies hazards and safety barriers, but does not consistently translate hydrogen PEMFC–LFP operation into observable competence assessment evidence and implementation pathways. This paper develops a demonstrator-anchored and regulatory-grounded competency framework for marine engineers operating compressed hydrogen PEMFC-lithium iron phosphate (LFP) battery–electric propulsion systems. A structured purposive narrative synthesis combined prototype vessel testing evidence with regulatory safety training, and competency framework literature. The experimental operational data, including compressed hydrogen supply, pressure regulation, PEMFC charging, battery buffering, propulsion current demand, voltage sag, state-of-charge response, monitoring tasks, alarms, and emergency isolation, were used as operational anchors rather than calibrated performance validation evidence. The analysis identified six competency domains. Compared with IGF/LNG model course training, the largest hydrogen-specific competence gaps concerned compressed hydrogen handling, PEMFC purge and shutdown logic, battery-buffered propulsion monitoring, integrated emergency shutdown, and communication during abnormal operation. These findings were translated into assessable learning outcomes, a provisional 40 h training module, instructor prerequisites, practical assessment evidence, a proposed digital twin/VR supplement, and a staged implementation roadmap. The proposed framework provides a structured pilot pathway. It translates operational testing evidence into assessable maritime education and training. It also establishes a foundation for future competency development and certification for commercial vessels. Full article

Railway transport management has changed dramatically over the past 50 years. The advent of computer technology and the capacity for information transmission brought greater safety and the ability to remotely control interlocking devices. These enable the centralisation of railway transport management, leading to higher operational efficiency and reduced staffing costs. At the same time, this technological progress has enabled the development of additional automation functions, which we can abbreviate as ARS (Automated Route Setting). The international designation Automatic Route Setting (ARS) includes actions that enable the automation tool to execute instructions to the signal box without the intervention of operating personnel (the dispatcher). Their importance increases with line speed and the size of the remotely controlled area. Thanks to them, the dispatcher gains time because the ARS can automatically resolve some operational situations or allow the dispatcher to address them in advance, thereby distributing the workload over a wider time window. However, the interlocking system itself remains the primary safety mechanism and will prevent ARS if any element of the infrastructure is occupied. At the same time, it is not possible to automate safety-critical functions that require direct assistance from the operating personnel. In the article, the authors analysed functions in which ARS is currently widely used. In the next part, they focused on the possible expansion of the palette of these functions that could be included in the ARS regime using multi-criteria analysis. The WSA method was applied using data obtained from routine users of the system. This approach enabled the incorporation of practical operational experience into the evaluation process and provided an empirical basis for assessing and prioritising the analysed functions. The next step was a safety-critical analysis and determination of the conditions under which they could be included in the ARS regime. The safety-critical functions are left aside. It is assumed that these will still have to be performed by the operator, not by the ARS. Detailed implementations and quantification of their impacts on the dispatcher’s activities are then carried out for selected ARS functions. The analysis therefore yields a prioritised ranking of ARS functions, indicating the order in which their implementation would be most appropriate from an operational perspective. This ranking provides a systematic basis for the phased deployment of ARS functionalities, considering their expected operational benefits and practical applicability in railway traffic management. The last part of the article is a look into the future, because the development in the field of safe communication between the train and the infrastructure (V2I) and the transmission of valid information provides many new challenges not only in the field of ARS itself, but also in the optimisation of the entire process of managing and organising rail transport. If we can use the ARS functions today, it is only a matter of technical development to be able, for example, to guide trains to the exact time when a train route will be built for this train. This will also enable optimising the train’s energy consumption and tracking capacity use. The ideal state is when the infrastructure fully communicates with the train in GoA4 mode and optimises both the train’s ride and the use of the infrastructure. Full article

Background: As construction safety training increasingly shifts toward virtual reality (VR)-based learning platforms, concerns have emerged regarding whether real ambient and virtual visual conditions may affect safety knowledge learning and cognitive responses; Methods: This study examined how real ambient factors (temperature and sound) and virtual visual factors (window view and visual complexity) influenced perceived environmental distraction, subjective workload, learning experience, and EEG-based cognitive responses in a VR classroom. Forty-eight university students completed a construction safety knowledge learning task and were categorized into high- and low-performing groups based on learning gains. Subjective responses were measured using post-experimental questionnaires, while EEG indicators included mental workload, attention, and mental fatigue. Results: Based on independent-samples t-tests, group-specific Spearman correlations, descriptive analyses, and two-way interaction regression, the results revealed: (1) Low-performing students reported greater disturbance from temperature and sound, and higher mental demand and effort. (2) Nominal associations between environmental factors and subjective outcomes were broader in the low-performing group. (3) The EEG results served mainly as supplementary descriptive evidence, showing individual variability in cognitive responses. (4) Learning scores varied across environmental combinations, with Temperature × Sound emerging as the only significant interaction. Conclusions: These findings guide cognitively supportive VR-based construction safety learning environments. Full article

Medical device-related pressure injuries (MDRPIs) represent a significant subset of hospital-acquired pressure injuries, yet nurses’ knowledge regarding their prevention remains inadequately characterized. MDRPIs are a critical yet often overlooked patient safety issue in acute-care settings. This study aimed to evaluate the knowledge levels of nurses regarding MDRPI risk factors in a hospital located in Eastern Türkiye. A descriptive cross-sectional study was conducted with 284 nurses working in intensive care, palliative care, surgical, and internal medicine units between October and December 2024. Data were gathered through a structured questionnaire assessing knowledge of risk factors, evidence-based prevention strategies, and device-specific clinical competencies. Statistical analysis included descriptive statistics, ANOVA, and multiple regression modeling. Results revealed moderate overall MDRPI knowledge levels with significant variation by clinical unit, years of experience, prior training exposure, and frequency of medical device use. Regression analysis identified prior education on pressure injury prevention as the strongest predictor of knowledge scores. Nurses in intensive care units demonstrated higher knowledge and prevention competency compared to those in palliative care and general medical-surgical units (p < 0.001). These findings underscore the need for targeted, device-specific educational interventions and standardized training protocols to enhance nurses’ competency in MDRPI prevention across diverse clinical settings. Strengthening in-service training protocols is essential to reduce preventable patient harm and improve clinical outcomes in regional healthcare facilities. Full article

Post-earthquake structural inspections are critical for public safety and recovery, yet traditional manual assessments are slow, hazardous, and resource-intensive. This paper proposes a novel system that integrates a quadrupedal robot with a deep learning (DL) vision model to rapidly detect structural cracks and damage in the aftermath of earthquakes. A Unitree Go2 quadruped robot, equipped with cameras and sensors, is paired with a YOLOv8 instance segmentation network for near-real-time crack detection and localization. The approach addresses key limitations of manual post-disaster inspections by enabling operator-supervised, near-real-time visual crack screening in hazardous or hard-to-reach areas. The YOLOv8 model is trained on a curated dataset of crack and damage images to support crack detection and segmentation performance, and its advanced segmentation capabilities allow precise delineation of damaged regions. The integrated system is validated on a laboratory-scale concrete–steel frame with simulated damage. Preliminary results demonstrate that the Unitree Go2 quadruped robot can navigate and inspect structural elements while the AI model identifies and segments cracks and surface damage under near-real-time laboratory operating conditions. This work highlights the potential of combining advanced legged robotics and state-of-the-art DL for structural health monitoring (SHM), offering a preliminary visual screening tool that can support operator awareness and help prioritize areas requiring expert structural inspection. Full article

Background/Objectives: Contemporary healthcare systems face compound challenges (including technological acceleration, demographic aging, rising chronic disease burden, and growing patient expectations) that demand models that are simultaneously efficient, high-quality, and person-centered. Despite a substantial body of research addressing organizational design, human capital management, and clinical competencies, these dimensions have largely been theorized in isolation. This study aims to construct and justify an integrated theoretical framework explaining how organizational models, human capital, and transversal competencies may jointly shape care quality, patient safety, and institutional sustainability in healthcare organizations. Methods: A narrative literature review was conducted, integrating contributions from business economics, healthcare management, organizational psychology, and nursing sciences. This design was selected for its suitability in synthesizing heterogeneous, multidisciplinary knowledge into a coherent conceptual framework, a purpose for which systematic meta-analytic approaches are not appropriate. Sources encompassed 79 references: peer-reviewed journals (PubMed, JSTOR, Google Scholar), institutional reports (WHO, OECD, European Commission, Joint Commission), normative standards (ISO 30414:2018), and Italian regulatory frameworks, spanning foundational twentieth-century contributions through the most recent literature (2025). Results: Four principal findings emerged: (1) healthcare organizations are evolving from rigid hierarchical structures toward flexible, value-based configurations, with the Value-Based Healthcare (VBHC) paradigm redirecting institutional attention from service volume to patient-meaningful outcomes per unit of cost; (2) transversal competencies (communication, empathy, emotional intelligence, teamwork, and transformational leadership) are closely associated with care quality and patient safety, with 70–80% of sentinel events associated with communication failures; (3) human capital, encompassing technical expertise and relational capacity, constitutes the primary lever of competitive advantage in healthcare institutions; and (4) the trajectory from pyramidal toward participatory and self-managed models is supported by international evidence, including the Buurtzorg experience in the Netherlands. Conclusions: The integrated three-pillar framework (combining resource-based theory and dynamic capabilities, Value-Based Healthcare, and evolutionary organizational theory) provides a theoretically grounded basis for understanding how organizational structure, human capital, and transversal competencies are jointly associated with clinical performance. Healthcare institutions should systematically integrate soft-skills training into professional education and invest in participatory organizational structures. Health policy should revise financing mechanisms to incentivize patient-meaningful outcomes over service volumes and support the broader transition toward Value-Based Healthcare models. The Italian SSN is discussed as an illustrative national context rather than as the primary empirical focus of the review. Full article

Road traffic accidents continue to be a major global cause of fatalities, disproportionately affecting pedestrians and other vulnerable road users. While deep reinforcement learning has proven effective in handling complex navigation tasks, providing formal stability and safety guarantees during both training and deployment remains a significant challenge. This paper introduces a dual-layer safety-aware framework for pedestrian avoidance in occluded urban driving. During training, a first-order Control Lyapunov–Barrier Function is integrated with Proximal Policy Optimization to promote goal-reaching stability and obstacle avoidance: the analytic Lie derivatives of the Lyapunov and barrier functions are embedded as a modifier in the advantage estimate, providing explicit stability and safety signals that accelerate convergence toward safe, goal-reaching behavior without disrupting the standard policy update. At deployment, a higher-order Control Lyapunov–Barrier Function, realized through a quadratic programming safety filter, acts as a safety shield that projects the nominal acceleration onto the intersection of the second-order Lyapunov and barrier feasibility sets; the barrier function is further extended with relative velocity terms to account for dynamic pedestrian motion. Experiments with a four-wheeled vehicle in the Webots simulator show that the framework reliably reaches the goal, avoids an occluded pedestrian across a range of crossing speeds, and improves task success rates and safety-constraint adherence relative to Proximal Policy Optimization and a conventional higher-order safety filter baseline, particularly during emergency braking maneuvers. Full article

Accurate modeling of degradation progression in rotating machinery remains challenging in real industrial systems, where data are inherently limited and imbalanced because of safety-critical operations and associated risks, and the cost of acquiring fault data is high. These conditions make it difficult for data-driven approaches to reliably capture the evolution of degradation over time. To address this challenge, this study proposes a hybrid framework that models degradation progression as a set of distinct behavioral regimes driven by loss of lubrication. The proposed framework first applies adaptive scaling guided by an $\alpha$ parameter derived from Root Mean Square (RMS) deviation of the vibration signals relative to the baseline condition, aiming to mitigate data leakage during preprocessing while improving robustness to data imbalance. It then performs baseline-referenced deep feature learning using a lightweight Long Short-Term Memory (LSTM) model trained only on baseline data. The trained model is subsequently used to encode the entire dataset into latent representations, which are finally clustered using Mini-Batch K-Means to organize distinct degradation-related behavioral regimes. Results on both real-world and experimental datasets demonstrate that the learned latent representations strongly agree with the degradation regimes associated with baseline characterization and $\alpha$-guided progression patterns, achieving an Adjusted Rand Index (ARI) of 1.0 across both datasets with respect to the internally defined reference stages. Full article

Wireless sensor networks underpin the Internet of Medical Things (IoMT), where connected medical devices and wearable body-area sensors stream patient telemetry across hospital networks, and securing this traffic is safety-critical. Machine learning intrusion-detection systems for the IoMT are usually evaluated within a single public dataset, where they report near-perfect detection scores; whether that accuracy predicts performance in a different hospital has not been measured systematically. We frame the deployed detector as a lightweight human-in-the-loop edge AI agent that observes local traffic, asks an analyst to label a small set of representative flows, trains a tiny model on-premises, and monitors the traffic distribution for drift. Using cross-corpus shift across four public datasets, comprising three sensor-network corpora (WUSTL-EHMS-2020, CIC-IoMT-2024, TON-IoT) and an unrelated enterprise-network corpus, as a reproducible stand-in for cross-hospital deployment, we find the shift is near its theoretical maximum on all twelve transfer directions, that unlabeled domain-adaptation methods collapse on the hardest medical-telemetry targets, and that external pretraining adds no measurable benefit once training schedules are matched, consistent with a domain-adaptation error bound that the measured divergence renders vacuous. The same 1206-parameter network trained from scratch on ten to fifty locally labeled flows matches or exceeds every transfer alternative across all four corpora; on the six sensor-network transfer directions it also outperforms larger and classical local models. A closed-loop evaluation on simulated drifting streams shows the calibrated trigger detects drift within two 500-flow windows at under one false alarm per hundred stationary windows, retraining restores accuracy, and the agent tolerates ten percent analyst error for about five F1 points; we also map the detector’s exposure to white-box evasion and targeted label poisoning. Full article

With the progressive opening of urban low-altitude airspace, drone applications increasingly rely on real-time environmental perception for safe navigation in complex, obstacle-dense environments. Traditional planning algorithms struggle to efficiently process spatially heterogeneous sensor data, leading to computational bottlenecks and unsafe trajectories. To bridge raw perception and agile decision-making, we propose an Environment-Aware Adaptive Heuristic Search (EA-AHS) framework that translates local obstacle density into adaptive heuristic weights via a sliding window, while a historical feedback loop enables macro-level parameter adaptation. Experiments demonstrate that EA-AHS reduces planning time by 87.9% in a simple maze and by 81.0% in a complex point-block map compared to standard A*. In a 3D urban scenario, EA-AHS trades a modest computational overhead for substantial safety gains: its cumulative risk is only 48% of that of standard A* and 57% of that of weighted A*, and its minimum obstacle distance increases to 6.04 m. Unlike learning-based methods, EA-AHS requires no pre-training or neural inference, offering a highly practical, lightweight solution for resource-constrained onboard sensor systems. Full article