Search Results (54)

Food allergies impose strict constraints on dietary decision-making, necessitating recommender systems that guarantee safety without compromising nutritional quality or user satisfaction. Existing systems often treat safety as a preference, failing to meet rigorous safety-critical standards or account for complex interactions between allergens, nutrition, and visual appeal. To address this issue, we propose a structured, multi-objective recipe recommendation framework. In this framework, rather than being modeled as an additive objective, allergen safety is prioritized through a safety-aware penalty-based ranking mechanism. The framework integrates three core modules: an allergen safety score accounting for cross-reactivity and cooking conditions; a nutritional balance score aligned with Dietary Reference Intakes (DRIs); and a neural-derived visual appeal score. In evaluation, we conducted a controlled user study ($N=20$) to evaluate the framework against single-factor baselines. Our integrated strategy consistently outperforms all single-factor baselines across evaluated metrics. Sensitivity analysis further confirms that safety-aware ranking ensures stable recommendation behavior across diverse preference profiles. Notably, behavioral analysis revealed a decision–action discrepancy, wherein users exhibited more risk-averse behavior during actual interactions than their explicitly stated preferences suggested. These findings suggest that prioritizing safety through safety-aware ranking mechanisms, together with multi-objective optimization, provides a robust foundation for personalized health-aware dietary support. Full article

To improve the reliability and path quality of three-dimensional UAV path planning in complex mountainous environments, this paper proposes an Adaptive Elite Differential Gold Rush Optimizer (AEDGRO). The main novelty of AEDGRO lies in the coordinated integration of three enhancement mechanisms into the original Gold Rush Optimizer: chaotic good-point initialization for improving initial population coverage, adaptive elite differential mining for strengthening exploitation around promising regions, and stagnation-aware Gaussian–Cauchy mutation for escaping local optima. A UAV path-planning model is constructed by considering path length, altitude fluctuation, trajectory smoothness, terrain collision avoidance, threat-region avoidance, and UAV safety clearance. The experimental results on the IEEE CEC2017 benchmark suite show that AEDGRO obtains the best Friedman average ranking of 1.63, outperforming the original GRO with a ranking of 4.80. In the UAV path-planning experiments, AEDGRO achieves the lowest mean fitness value of 235.69 and the smallest standard deviation of 7.55, indicating better path quality and stronger robustness than the compared algorithms. The generated trajectories are smoother and can effectively avoid mountainous terrain and threat regions. These results demonstrate that AEDGRO has clear advantages in global optimization accuracy, convergence stability, and UAV path-planning applicability. Full article

Evaluating conditional time-series generation models remains challenging in battery research, where degradation data are often limited and experiments cover only a small number of operating conditions. The widely used Fréchet Inception Distance (FID) summarizes all conditions into a single score, which can obscure failures under rare but safety-critical conditions. Several condition-aware extensions of FID, including Conditional Fréchet Inception Distance (CFID), partially address this limitation by evaluating each condition separately. However, these approaches do not assess whether physically meaningful relationships between operating conditions are preserved, and their reliability deteriorates when only a few samples are available for each condition. To address these issues, we propose a three-layer diagnostic framework for evaluating conditional generative models under limited-data conditions. The first layer, Stratified Fréchet Distance, identifies the specific operating conditions and degradation phases where generation quality degrades. The second layer, based on Conditional Response Consistency (CRC), Conditional Distance Ratio (CDR), and Mean-Order Preservation (MOP), evaluates whether the model preserves the distance structure and ordering between conditions. MOP detects condition-ordering defects that CRC cannot identify when the real data distance matrix is non-monotone. This layer also enables statistically meaningful comparisons even when only a small number of samples are available. The third layer detects strata where statistical estimates are unreliable and provides a more stable alternative for evaluation. We validate the framework on four battery degradation datasets using two generative model architectures. The proposed approach reveals condition-specific failures that are not captured by conventional FID. It localizes generation errors to the late-stage high-temperature degradation regime that is most relevant to battery safety. The framework also detects structural distortions with statistical significance. In addition, it consistently ranks physics-informed model variants across quality differences spanning seven orders of magnitude. These results demonstrate that the proposed framework provides a practical and physically interpretable evaluation methodology for conditional generative modeling in battery degradation analysis. Full article

Scaffold-related falls remain a major safety concern in demolition projects, where temporary access systems are frequently erected, modified, used, and dismantled under changing structural and site conditions. These characteristics complicate risk prioritization because scaffold failures may involve interacting human, technical, organizational, and environmental factors. This study develops an expert-based risk prioritization framework for scaffold-related fall risks in demolition projects by integrating Failure Mode and Effects Analysis (FMEA), interval neutrosophic fuzzy (INF) theory, Decision-Making Trial and Evaluation Laboratory (DEMATEL), and Multi-Attributive Border Approximation Area Comparison (MABAC). Using the 4M1E perspective, namely Man, Machine, Material, Method, and Environment, 37 demolition-specific failure modes were identified through literature review and expert elicitation. Ten experts evaluated these failure modes using the SODE criteria, namely Severity, Occurrence, Detection difficulty, and Expected Cost impact. INF theory was used to represent uncertainty, hesitation, and judgmental variation in expert assessments. INF-DEMATEL was applied to examine interrelationships among the SODE criteria and derive interdependence-aware criterion weights, while INF-MABAC was used to rank the failure modes according to their distance from the Border Approximation Area. The framework was illustrated through an empirical application in Taiwan’s demolition industry. The results identified Severity as the most influential criterion. The highest-priority failure modes were insufficient safety awareness, improper scaffold-to-structure anchoring, and inadequate scaffold maintenance and inspection governance. Comparison with risk priority number (RPN)-based methods and sensitivity analyses using expert exclusion and Severity-weight variation showed that the ranking was generally consistent and reasonably stable under the tested conditions. The proposed framework provides a structured, uncertainty-aware decision-support procedure for identifying prevention priorities in demolition scaffold operations. Full article

Accurate meteorological visibility estimation is vital for transportation safety and environmental monitoring. However, modeling the inherent nonlinear spatial and spectral degradations in hazy environments remains challenging. While recent Large Vision-Language Models (LVLMs) offer strong scene understanding, they lack the regression precision required for visibility estimation. In this paper, we propose the Visibility-Aware Refined CNN (VISR-CNN), a dual-stream architecture that synthesizes local spatial cues with global frequency-domain signatures. The model integrates a Multi-Scale Transmission Attention (MSTA) module, which uses parallel dilated convolutions to estimate atmospheric transmission, and a Global Frequency Branch that utilizes 2D Real Fast Fourier Transforms (RFFT) with Spectral Gating to quantify visibility-dependent blurring. A progressive training strategy is introduced to decouple spectral and spatial optimization, and a physics-informed loss function is designed to supervise numerical regression while enforcing a monotonic ranking constraint consistent with physical light-attenuation laws. Results on the HKCHC-VD dataset show that VISR-CNN achieves state-of-the-art performance (MAE: 1.54 km; RMSE: 2.31 km), representing a 13.0% improvement over VisNet. Further evaluations on the CP1 and SWH datasets confirm robust generalization, reducing overall MAE by 21% and 20%, respectively, compared with the hybrid ResNeXt-50 + ViT model. Notably, in safety-critical range (0–10 km), VISR-CNN reduces RMSE for the HKCHC-VD, CP1, and SWH datasets by approximately 55%, 64%, and 71%, respectively, when compared with VisNet. These findings demonstrate the superiority of specialized, physics-grounded architectures over general-purpose LVLMs for high-precision meteorological regression. Full article

As Industry 4.0 transforms manufacturing operations, maintenance organizations face a group decision-making problem: how to consolidate diverse expert judgments into a defensible, transparent ranking of the competencies that maintenance personnel most need. This paper applies an integrated Delphi–AHP framework—with explicit notation, operators, and diagnostics—to prioritize maintenance competencies in advanced-manufacturing settings. The Delphi stage consolidates expert-generated items under median–interquartile-range consensus and round-to-round stability rules, while the Analytic Hierarchy Process (AHP) transforms validated pairwise comparisons into ratio-scale priority weights through geometric-mean Aggregation of Individual Judgments (AIJ) and eigenvector derivation. Consistency screening (CI/CR), inter-rater agreement (Kendall’s W), and perturbation-based sensitivity analysis accompany the resulting weight vector. A bounded AI-assisted consistency-check step supports terminology harmonization during Delphi statement consolidation, subject to explicit human-validation constraints. A panel of fifteen industry experts participated in the study; five competency dimensions and twenty-nine indicators were retained through three Delphi rounds. AHP weighting identified Basic Knowledge and Skills as the highest-priority dimension, followed by Safety and Regulation Awareness and Problem-Solving Ability. Aggregated pairwise comparison matrices, local and global weights, and sensitivity results are reported to support reproducibility. The study contributes a rigorously specified application of combined Delphi–AHP to a domain—Industry 4.0 maintenance asset management—where multi-criteria decision analysis has seen limited formal application, and closes common specification gaps in published Delphi–AHP implementations. Full article

Recent advancements in autonomous driving technology are transforming the automotive industry, with advanced driver assistance systems (ADAS) recognized as a crucial transitional technology toward fully autonomous driving. ADAS enhances driver safety and comfort through features such as emergency braking, lane-keeping, and adaptive cruise control, ultimately aiding in traffic accident prevention and reduction in driver fatigue. However, commercial ADAS implementations show substantial variability due to differences in sensor configurations, operational design domain (ODD) definitions, and operational criteria across automakers. To address this gap, this study provides a structured comparative review of commercialized ADAS technologies across 11 major Western and Asian automakers. By encompassing both Western and Asian OEMs, this study compares manufacturer-declared sensor configurations, ODD settings, activation conditions, driver-monitoring requirements, takeover and fallback logic, and update-related characteristics. The review identifies implementation-level differences that affect comparability, user understanding, validation requirements, and standardization needs. Rather than ranking OEM systems by safety performance, this study clarifies the trade-offs among redundancy-oriented, camera-centric, HD-map-dependent, geofenced, and OTA-driven ADAS strategies. The findings support future work on standardized ODD communication, user-centered HMI design, independent validation, and update-aware review frameworks for commercial ADAS. Full article

Computer Numerical Control (CNC) machining centers are critical assets in discrete manufacturing, yet many shop floors still rely on periodic expert judgment for machine selection and workload allocation. This practice is unsuitable for high-mix production because machine condition and risk can change rapidly due to tool wear, thermal drift, coolant variation, and alarms. Moreover, decision evidence is fragmented and often incomplete across controller and programmable logic controller signals, production records, and inspection results, making manual evaluation time-consuming and prone to misjudgment. Static rankings can also break down under unforeseen shop-floor disruptions, requiring rapid event-driven re-prioritization and rescheduling. To address these challenges, this research proposes a shop-floor decision intelligence pipeline that executes a rolling-window, uncertainty-aware ranking-and-dispatch loop directly on the shop floor. The industrial compute node continuously collects multi-source operational evidence, normalizes it into a unified event representation, and aggregates rolling-window indicators for each machine. A mapping structure then converts these indicators into neutrosophic triplets that separate performance from evidence credibility. Using this representation, a shop-floor decision procedure continuously updates machine priority scores using a TOPSIS procedure, which are further translated into workload allocation and persistence-confirmed protective action requests. A case study demonstrates end-to-end operation. It shows that the top-ranked machines remain stable under risk-aversion and weight-uncertainty analyses, while the protective logic prevents unsafe dispatching when reject-level conditions persist under reliable evidence. Overall, the proposed pipeline reframes CNC machine selection as a rolling-window, evidence-driven decision process and provides a pathway toward near-real-time and safety-aware shop-floor coordination. Full article

Time headway is a key parameter for describing car-following behavior and microscopic traffic flow characteristics, and it is important for traffic safety analysis, road design, and optimizing intelligent-driving strategies. Existing research offers limited insight into the heterogeneity of time headway under different vehicle types and lane conditions. It is particularly important to investigate how time headway distributions differ across lane–vehicle-type combinations on highways, as these differences can affect safety evaluation and operational performance. This study is based on drone-captured vehicle trajectories from the publicly available HighD dataset. We select 378,751 vehicle–frame trajectory records; these records are used to construct valid follower–leader pairs and derive time headway (THW) samples for distribution fitting. Eight subsets are formed by combining two lane positions (inner vs. outer) and four follower–leader vehicle-type pairs (car–car, car–truck, truck–car, truck–truck). Six candidate distributions (Lognormal, Log-logistic, Burr, Weibull, Gamma, and Logistic) are fitted using maximum likelihood estimation, and their fit is evaluated using Kolmogorov–Smirnov, Anderson–Darling, and Chi-square tests, which are fused via an entropy-weighted composite score for model ranking. Results show pronounced heterogeneity across lane–vehicle-type subsets: Inner-lane samples exhibit smaller and more concentrated time gaps, whereas outer-lane samples show larger mean gaps, stronger dispersion, and heavier upper tails. Overall, Lognormal(3P) is selected as the top-ranked model in 5 of 8 subsets (62.5%), while Burr(4P) (car–truck, outer lane), Gamma(3P) (truck–car, outer lane), and Weibull(3P) (truck–truck, inner lane) are optimal in the remaining subsets. These findings indicate that lane position and vehicle-type pairing materially affect THW distributional characteristics, providing quantitative guidance for lane- and vehicle-aware traffic modeling, safety-oriented assessment, and intelligent-driving strategy design. Full article

Background: Parents serve as the primary decision-makers for childhood vaccination while also making decisions regarding their own vaccination, yet vaccination decision drivers are typically studied separately by vaccine type or target population. Methods: This study investigated parental decision-making processes for two self-paid and non-National Immunization Program vaccines in China, childhood rotavirus vaccine and adult influenza vaccine, by utilizing a structured survey grounded in the World Health Organization Behavioral and Social Drivers of Vaccination framework. Spearman’s rank correlation coefficients were used to assess the consistency of parental attitudes toward the two vaccines across behavioral and social driver domains. Structural equation models were conducted separately for childhood and adult vaccines to examine decision-making pathways. Results: The findings indicated that parental drivers related to awareness, social processes, and practical issues showed a high consistency across adult and childhood vaccination decisions (r > 0.7), whereas the consistency in vaccination behaviors remained low (r = 0.21). Compared with adult vaccination, childhood vaccination decisions were more strongly influenced by vaccine safety concerns and healthcare practitioners’ recommendations, which emerged as key drivers. Furthermore, family norms emerged as an effectively shared driver of vaccination decisions for both adult and childhood vaccines (adult: β = 0.784; childhood: β = 0.970). Conclusions: By jointly synthesizing adult and childhood vaccination decisions from a parental perspective, this study provides crucial evidence to support the development of integrated, family-centered strategies to improve vaccine uptake. Full article

Autonomous ground vehicles operating in structured and semi-structured environments—such as urban roads, parking lots, and logistics warehouses—require fast, reliable, and collision-free path planning on occupancy grid maps. Existing metaheuristic planners often suffer from premature convergence, insufficient population diversity, and poor feasibility maintenance, limiting their deployment in safety-critical vehicular navigation. This paper proposes a multi-strategy enhanced Crested Porcupine Optimizer (MSCPO) that systematically addresses these limitations through four coordinated enhancements: chaos-opposition initialization with feasibility repair to ensure high-quality and diverse initial routes; a diversity-coupled adaptive mechanism for dynamic strategy scheduling throughout the search; elite-guided differential Lévy perturbation to escape local optima and accelerate convergence; and a two-stage safety-aware objective with elite local refinement to sharpen final solution precision. Experiments on four representative grid maps with varying obstacle densities, conducted over 30 independent runs per algorithm, demonstrate that MSCPO consistently outperforms state-of-the-art metaheuristic planners and deterministic baselines in path length, smoothness, and convergence speed. Statistical analysis via Wilcoxon rank-sum and Friedman tests confirms the significance of the improvements. An ablation study quantifies the individual contribution of each enhancement module, confirming the practical effectiveness of MSCPO for autonomous vehicle navigation tasks. Full article

In-service deviations of viscous dampers can reduce the collapse safety margin of viscous-damped structures under strong earthquakes. This study examines two representative mechanisms: global degradation of the damper group and local failure of a subset of dampers. Incremental dynamic analyses are conducted for five damper-state scenarios using the 22 far-field ground-motion records recommended by ATC-63. To support reliability-oriented, uncertainty-aware collapse-capacity comparison with limited records, three complementary probabilistic inference frameworks are developed: an event-based fragility model using binary collapse indicators, a drift-margin model leveraging continuous deformation information from non-collapse responses, and a fusion model that combines both sources via a weighted composite likelihood with fusion strength governed by the weight w. For each scenario, the capacity scale parameter μ_m is reported as IM_50,m, and record-level bootstrap resampling is used to construct interval estimates. Multi-scenario effects are further summarized by the ensemble mean reduction b and inter-path dispersion σ_damper, offering compact measures of systematic shift and pathway-to-pathway variability. Results indicate a dominant systematic downward shift in median collapse capacity, with IM_50,m reduced by approximately 2.4–2.9% overall, whereas differences among degradation pathways are secondary and bounded by the intervals. Scenario rankings remain consistent across the three frameworks; fusion outputs show weak sensitivity to w and yield tighter interval constraints on σ_damper than the event-only baseline. The resulting interval-based parameters enable risk- and reliability-informed interpretation of degradation effects and provide a consistent basis for uncertainty quantification in probabilistic performance comparisons across scenarios. Full article

Long-term data loss resulting from sensor malfunctions, communication interruptions, and other factors in Structural Health Monitoring (SHM) significantly undermines the reliability of damage identification and safety assessment. Existing methods—ranging from statistical approaches and low-rank matrix completion to traditional machine learning and deep learning imputation techniques—often suffer from either limited accuracy or excessive model size and slow inference, making deployment in resource-constrained scenarios difficult. To address these challenges, this paper proposes TEFN–Imputation, a lightweight and efficient time-series imputation model. This model utilizes observation-driven non-stationary normalization to mitigate the impact of time-varying characteristics and dimensional discrepancies. It employs linear projection for temporal length alignment and constructs BPA-style mass representations from dual perspectives of time and channel. Furthermore, it replaces strict Dempster–Shafer belief combination with an expectation-based evidential aggregation (readout), thereby significantly reducing computational overhead while enabling uncertainty-aware evidential indicators for interpretation rather than claiming a direct accuracy gain from uncertainty modeling. The observed accuracy and robustness improvements are primarily attributed to the normalization and dual temporal–channel modeling design under the same lightweight readout. Systematic experiments on two real-world bridge monitoring datasets, Z24 and Hell Bridge, demonstrate that TEFN consistently maintains low Mean Absolute Error (MAE) and minimal volatility across various combinations of training and testing missing rates, exhibiting high robustness against variations in missing rates and train–test mismatches. Concurrently, compared to RNN and large-scale Transformer baselines, TEFN reduces parameter count and CPU inference time by one to two orders of magnitude. Thus, it achieves a superior trade-off among accuracy, efficiency, and model scale, making it highly suitable for online SHM and imputation tasks in practical engineering applications. Across the settings on Z24, TEFN achieves a mean MAE of 0.218 with a standard deviation of 0.002, while using only 0.02 MB parameters and 2.73 ms per batch CPU inference. Full article

Tracking worker identities across cameras is increasingly important for advanced construction site monitoring, such as safety and productivity monitoring. However, current computer vision-based tracking faces challenges in reliably associating worker identities due to frequent occlusions and extreme viewpoint shifts between aerial and ground cameras, resulting in fragmented trajectories and ID switches. This study proposes a Dual-Stage Graph-based Association framework that integrates worker detections across multiple views using complementary Re-identification models and camera-aware adaptive thresholding. The framework synergistically combines TransReID for viewpoint-invariant global features and BPBReID for occlusion-robust part-based features, producing more discriminative representations. Data association leverages a graph-based clustering approach to combine representation features, camera topology, and temporal cues for robust identity maintenance. The first stage enables cross-view clustering while preventing false matches, and the second stage ensures long-term identity stability through EMA-based gallery management. Experiments on two construction sites demonstrate that the proposed framework achieves an HOTA of 39.85% and an IDF1 of 63.58%, outperforming existing baselines while reducing ID switches by 35.0%. Results on the AG-ReID.v2 benchmark demonstrate strong generalization with 90.82% Rank-1 accuracy in aerial-to-CCTV matching. The approach highlights initial feasibility for cross-view multi-camera tracking in construction with potential for extension to more complex industrial environments. Full article

The reliability of photovoltaic (PV) equipment is vital for ensuring the safe and stable operation of power systems. While multimodal large language models (MLLMs) open up promising avenues for intelligent fault diagnosis, they often falter when confronted with the heterogeneity of PV data—where visual observations come from different sensor modalities (e.g., visible, infrared, and thermal) and display strong domain-dependent variations. Conventional Low-Rank Adaptation (LoRA) is not expressive enough to model such modality-aware differences, which can result in insufficient exploitation of informative patterns. To overcome this limitation, we propose PV-FaultExpert, a domain-adaptive MLLM designed specifically for PV equipment fault analysis. PV-FaultExpert is built upon DyLoRA (Dynamic Expert Routing with LoRA), a dynamic routing strategy that reformulates standard LoRA into a shared low-rank component coupled with multiple expert-specific adapters. A routing module then selects expert paths according to input characteristics, allowing the model to adapt to diverse modalities while maintaining parameter efficiency. Moreover, we construct a PVfault diagnosis dataset via ChatGPT-4o-assisted chain-of-thought reasoning and subsequent expert verification, which both supports model training and enables rigorous evaluation of our method. Extensive experiments demonstrate that PV-FaultExpert consistently surpasses strong baselines, including GPT-4 and Claude-3, across multiple evaluation criteria, producing fault analysis reports that are accurate, interpretable, and aligned with safety-critical requirements. Full article