Safety

16 pages, 274 KB

Open AccessArticle

Two Wheels or Four Wheels: A Comparative Study of Police Tasks on Bicycle vs. Car in Saguenay

by Pier-Luc Langlais, Marc-Antoine Masse and Martin Lavallière

Safety 2026, 12(1), 8; https://doi.org/10.3390/safety12010008 - 9 Jan 2026

Modern police work requires a high degree of versatility, shifting between sedentary tasks and intense physical demands. While bicycle patrols are recognized as a tool for enhancing community policing, few empirical studies have examined the specific nature and frequency of the tasks performed [...] Read more.

Modern police work requires a high degree of versatility, shifting between sedentary tasks and intense physical demands. While bicycle patrols are recognized as a tool for enhancing community policing, few empirical studies have examined the specific nature and frequency of the tasks performed by bicycle patrol officers. This study aims to compare the professional tasks of bicycle and car patrol officers in the city of Saguenay, Québec, over a three-year period. A retrospective analysis of 539 computer-aided dispatch (PCAD) entries was conducted for eight male officers (six on bicycles, two in police cars) during the summer months of 2021 to 2023. We analyzed task frequency, duration, priority, and risk level using descriptive statistics. Results showed that while both patrol types performed similar core tasks, such as citizen assistance, enforcement of municipal regulations, and responses to suspicious individuals, bicycle patrols were associated with significantly longer total PCAD-recorded intervention times (49 ± 47 min vs. 33 ± 29 min). Moreover, the distribution of call types suggests a slightly higher proportion of interventions occurring in public spaces or involving direct citizen contact, although this does not constitute a measure of increased proximity. No significant differences were observed in terms of priority or risk. Because the PCAD system does not systematically record on-scene time, the longer durations observed for bicycle patrols cannot be interpreted as qualitative advantages. Instead, the study reveals operational similarities alongside noteworthy differences between patrol types. As one of the first Canadian CAD-based analyses of bicycle patrol tasks, this research underscores the need for future studies capable of isolating on-scene time and examining the qualitative dimensions of police–citizen interactions. Full article

43 pages, 701 KB

Open AccessReview

New Trends in the Use of Artificial Intelligence and Natural Language Processing for Occupational Risks Prevention

by Natalia Orviz-Martínez, Efrén Pérez-Santín and José Ignacio López-Sánchez

Safety 2026, 12(1), 7; https://doi.org/10.3390/safety12010007 - 8 Jan 2026

Abstract

In an increasingly technologized and automated world, workplace safety and health remain a major global challenge. After decades of regulatory frameworks and substantial technical and organizational advances, the expanding interaction between humans and machines and the growing complexity of work systems are gaining [...] Read more.

In an increasingly technologized and automated world, workplace safety and health remain a major global challenge. After decades of regulatory frameworks and substantial technical and organizational advances, the expanding interaction between humans and machines and the growing complexity of work systems are gaining importance. In parallel, the digitalization of Industry 4.0/5.0 is generating unprecedented volumes of safety-relevant data and new opportunities to move from reactive analysis to proactive, data-driven prevention. This review maps how artificial intelligence (AI), with a specific focus on natural language processing (NLP) and large language models (LLMs), is being applied to occupational risk prevention across sectors. A structured search of the Web of Science Core Collection (2013–October 2025), combined OSH-related terms with AI, NLP and LLM terms. After screening and full-text assessment, 123 studies were discussed. Early work relied on text mining and traditional machine learning to classify accident types and causes, extract risk factors and support incident analysis from free-text narratives. More recent contributions use deep learning to predict injury severity, potential serious injuries and fatalities (PSIF) and field risk control program (FRCP) levels and to fuse textual data with process, environmental and sensor information in multi-source risk models. The latest wave of studies deploys LLMs, retrieval-augmented generation and vision–language architectures to generate task-specific safety guidance, support accident investigation, map occupations and job tasks and monitor personal protective equipment (PPE) compliance. Together, these developments show that AI-, NLP- and LLM-based systems can exploit unstructured OSH information to provide more granular, timely and predictive safety insights. However, the field is still constrained by data quality and bias, limited external validation, opacity, hallucinations and emerging regulatory and ethical requirements. In conclusion, this review positions AI and LLMs as tools to support human decision-making in OSH and outlines a research agenda centered on high-quality datasets and rigorous evaluation of fairness, robustness, explainability and governance. Full article

(This article belongs to the Special Issue Advances in Ergonomics and Safety)

► Show Figures

Figure 1

23 pages, 3238 KB

Open AccessArticle

Agricultural Injury Severity Prediction Using Integrated Data-Driven Analysis: Global Versus Local Explainability Using SHAP

by Omer Mermer, Yanan Liu, Charles A. Jennissen, Milan Sonka and Ibrahim Demir

Safety 2026, 12(1), 6; https://doi.org/10.3390/safety12010006 - 8 Jan 2026

Abstract

Despite the agricultural sector’s consistently high injury rates, formal reporting is often limited, leading to sparse national datasets that hinder effective safety interventions. To address this, our study introduces a comprehensive framework leveraging advanced ensemble machine learning (ML) models to predict and interpret [...] Read more.

Despite the agricultural sector’s consistently high injury rates, formal reporting is often limited, leading to sparse national datasets that hinder effective safety interventions. To address this, our study introduces a comprehensive framework leveraging advanced ensemble machine learning (ML) models to predict and interpret the severity of agricultural injuries. We use a unique, manually curated dataset of over 2400 agricultural incidents from AgInjuryNews, a public repository of news reports detailing incidents across the United States. We evaluated six ensemble models, including Gradient Boosting (GB), eXtreme Grading Boosting (XGB), Light Gradient Boosting Machine (LightGBM), Adaptive Boosting (AdaBoost), Histogram-based Gradient Boosting Regression Trees (HistGBRT), and Random Forest (RF), for their accuracy in classifying injury outcomes as fatal or non-fatal. A key contribution of our work is the novel integration of explainable artificial intelligence (XAI), specifically SHapley Additive exPlanations (SHAP), to overcome the “black-box” nature of complex ensemble models. The models demonstrated strong predictive performance, with most achieving an accuracy of approximately 0.71 and an F1-score of 0.81. Through global SHAP analysis, we identified key factors influencing injury severity across the dataset, such as the presence of helmet use, victim age, and the type of injury agent. Additionally, our application of local SHAP analysis revealed how specific variables like location and the victim’s role can have varying impacts depending on the context of the incident. These findings provide actionable, context-aware insights for developing targeted policy and safety interventions for a range of stakeholders, from first responders to policymakers, offering a powerful tool for a more proactive approach to agricultural safety. Full article

(This article belongs to the Special Issue Farm Safety, 2nd Edition)

► Show Figures

Figure 1

28 pages, 3549 KB

Open AccessArticle

Identification of Key Contributing Factors and Risk Propagation Paths in Safety Accidents at Chinese Chemical Enterprises

by Zhiheng Ni, Zhen Li, Mingyu Zhang and Otsile Morake

Safety 2026, 12(1), 5; https://doi.org/10.3390/safety12010005 - 5 Jan 2026

Abstract

To address the complex and uncertain causes of safety accidents in chemical enterprises, this study applied text mining techniques to systematically extract 29 causative factors from 422 accident reports. These factors were classified into five categories: personnel issues, resource management deficiencies, adverse organizational [...] Read more.

To address the complex and uncertain causes of safety accidents in chemical enterprises, this study applied text mining techniques to systematically extract 29 causative factors from 422 accident reports. These factors were classified into five categories: personnel issues, resource management deficiencies, adverse organizational atmosphere, organizational process flaws, and inadequate supervision. Based on the extracted factors, a complex network model of accident causation was constructed. Using degree centrality, betweenness centrality, and eigenvector centrality, seven core causative factors were identified, along with multiple peripheral factors closely linked to them. Bayesian network-based sensitivity analysis further revealed the factors that exert the greatest influence on accident occurrence, and subsequent path analysis uncovered several critical accident propagation paths. The findings reveal core causative factors and critical propagation paths, which may inform the prioritization of risk control measures under conditions of limited resources. Full article

► Show Figures

Figure 1

10 pages, 1866 KB

Open AccessArticle

Crowding, Risk, and Visitor Use Management on the Angels Landing Trail in Zion National Park

by Jeffrey N. Rose

Safety 2026, 12(1), 4; https://doi.org/10.3390/safety12010004 - 5 Jan 2026

Abstract

Zion National Park has seen substantial increased visitor use in recent years, bringing forward a number of visitor use management challenges. Many visitors consider the park’s Angels Landing trail, a steep and relatively challenging hike, a primary destination in the park. A number [...] Read more.

Zion National Park has seen substantial increased visitor use in recent years, bringing forward a number of visitor use management challenges. Many visitors consider the park’s Angels Landing trail, a steep and relatively challenging hike, a primary destination in the park. A number of well documented fatalities have been associated with the Angels Landing trail, prompting substantial risk management concerns. In the context of increased visitor use and increased attention to these fatalities, this research reviews literature on crowding and risk management before using National Park Service and media reports concerning 16 deaths associated with Angels Landing to characterize trends among age, gender, time of day, specific location, and other factors. Findings note that few of the fatalities occurred on the trail itself; those that did were not on the sections of the trail where risk management interventions have been installed, and none were associated with crowding or high visitor use. From these analyses, managers should consider disentangling notions of crowding and risk, particularly in light of new management strategies concerning permitting and limiting hikers on Angels Landing. Full article

► Show Figures

Figure 1

17 pages, 2210 KB

Open AccessArticle

The Use of a Device to Improve the Evacuation Performance of Hospitalized Non-Self-Sufficient Patients in Healthcare Facilities

by Simone Accorsi, Francesco Ottaviani, Aurora Fabiano and Dimitri Sossai

Safety 2026, 12(1), 3; https://doi.org/10.3390/safety12010003 - 24 Dec 2025

Abstract

Background: Fire emergency management in healthcare facilities represents a complex challenge, particularly in historic buildings subject to architectural preservation constraints, where progressive horizontal evacuation is objectively difficult. This study analyzes the effectiveness of an evacuation sheet employed by Hospital Policlinico San Martino to [...] Read more.

Background: Fire emergency management in healthcare facilities represents a complex challenge, particularly in historic buildings subject to architectural preservation constraints, where progressive horizontal evacuation is objectively difficult. This study analyzes the effectiveness of an evacuation sheet employed by Hospital Policlinico San Martino to improve the speed of evacuating non-self-sufficient patients in these buildings. Methods: This study involved evacuation simulations in wards previously selected based on structural characteristics. Healthcare personnel (male and female, aged between 30 and 55 years) conducted both horizontal and vertical patient evacuation drills, comparing the performance of the S-CAPEPOD^® Evacuation Sheet (Standard Model) with the conventional method (hospital bed plus and rescue sheet). This study focused on the night shift to evaluate the most critical scenario in terms of human resources. Results: The use of the evacuation sheet proved more efficient than the conventional method throughout the entire evacuation route, especially during the first 15 min of the emergency (the most critical period). Indeed, with an equal number of available personnel, the evacuation sheet enabled an average improvement of 50% in the number of patients evacuated. Conclusions: The data support the effectiveness of the device, confirming the theoretical premise that the introduction of the evacuation sheet—also due to its ease of use—can be an improvement measure for the evacuation performance of non-self-sufficient patients, despite limitations related to structural variability and the simulated nature of the trials. Full article

► Show Figures

Figure 1

23 pages, 1919 KB

Open AccessArticle

Machine Learning Assessment of Crash Severity in ADS and ADAS-L2 Involved Crashes with NHTSA Data

by Nasim Samadi, Ramina Javid, Sanam Ziaei Ansaroudi, Neda Dehestanimonfared, Mojtaba Naseri and Mansoureh Jeihani

Safety 2026, 12(1), 2; https://doi.org/10.3390/safety12010002 - 23 Dec 2025

Abstract

As the deployment of Automated Driving Systems (ADS) and Advanced Driver Assistance Systems (ADAS-L2) expands, understanding their real-world safety performance becomes essential. This study examines the severity and contributing factors of crashes involving vehicles equipped with ADS and ADAS-L2 technologies using NHTSA data. [...] Read more.

As the deployment of Automated Driving Systems (ADS) and Advanced Driver Assistance Systems (ADAS-L2) expands, understanding their real-world safety performance becomes essential. This study examines the severity and contributing factors of crashes involving vehicles equipped with ADS and ADAS-L2 technologies using NHTSA data. Using machine learning models on crash datasets from 2021 to 2024, this research identifies patterns and risk factors influencing injury outcomes. After data preprocessing and handling missing values for severity classification, four models were trained: logistic regression, random forest, SVM, and XGBoost. XGBoost outperformed the others for both ADS and ADAS-L2, achieving the highest accuracy and recall. Variable importance analysis showed that for ADS crashes, interactions with other road users and poor lighting were the strongest predictors of injury severity, while for ADAS-L2 crashes, fixed object collisions and low light conditions were most influential. From a policy and engineering perspective, this study highlights the need for standardized crash reporting and improved ADS object detection and pedestrian response. It also emphasizes effective human–machine interface design and driver training for partial automation. Unlike previous research, this study conducts comparative model-based evaluations of both ADS and ADAS-L2 using recent crash reports to inform safety standards and policy frameworks. Full article

► Show Figures

Figure 1

21 pages, 1701 KB

Open AccessArticle

Uncovering the Ergonomic Risks Threatening the Health of Underground Female Coal Mineworkers

by Ouma S. Mokwena, Thabiso J. Morodi and Joyce Shirinde

Safety 2026, 12(1), 1; https://doi.org/10.3390/safety12010001 - 19 Dec 2025

Abstract

Women in mining face unique health and safety challenges due to anatomical and physiological differences, making the assessment and management of ergonomic risks in underground coal mines critical. This study examines the ergonomic experiences of female mineworkers through six focus-group discussions, each comprising [...] Read more.

Women in mining face unique health and safety challenges due to anatomical and physiological differences, making the assessment and management of ergonomic risks in underground coal mines critical. This study examines the ergonomic experiences of female mineworkers through six focus-group discussions, each comprising eight participants, using a qualitative research design involving women actively engaged in core mining activities at three South African mines. Findings reveal that mining equipment and work environments often fail to accommodate the physiological needs of female workers, exposing them to a range of ergonomic hazards. Beyond physical risks, the study highlights organizational and systemic shortcomings, including inadequate implementation of existing policies and regulations. Poor hygiene in toilet facilities was also reported, with three out of eight participants taking medication for urinary tract infections, underscoring gaps in occupational health provision. The findings emphasize the urgent need for mine-specific ergonomic programs developed through participatory approaches, as part of a broader strategy to prevent musculoskeletal injuries and improve working conditions for female mineworkers. The establishment of the Women in Mining Forum further indicates that the industry is not yet fully prepared to support women in underground mining, highlighting the need for targeted interventions to create a safer, more inclusive work environment. Full article

► Show Figures

Figure 1

17 pages, 1011 KB

Open AccessArticle

Vulnerable Road Users in Romania: Forensic Autopsy-Based Analysis of Child and Elderly Fatalities

by Ştefania Ungureanu, Camelia-Oana Mureșan, Alexandra Enache, Emanuela Stan, Raluca Dumache, Octavia Vița, Ecaterina Dăescu, Alina-Cristina Barb and Veronica Ciocan

Safety 2025, 11(4), 125; https://doi.org/10.3390/safety11040125 - 15 Dec 2025

Abstract

Background: Vulnerable road users (VRUs), including children and older adults, face a high risk of fatal road traffic accidents (RTAs) due to limited protection and greater injury susceptibility. Romania reports some of the highest child and elderly RTA mortality rates in the European [...] Read more.

Background: Vulnerable road users (VRUs), including children and older adults, face a high risk of fatal road traffic accidents (RTAs) due to limited protection and greater injury susceptibility. Romania reports some of the highest child and elderly RTA mortality rates in the European Union. This study analyzed medico-legal autopsies from the Timisoara Institute of Legal Medicine (TILM) between 2017 and 2021 to compare fatalities in these two groups and identify key risk factors. Methods: A retrospective analysis was conducted on autopsy records of children (0–17 years) and older adults (>70 years) who died in RTAs during the study period. Data on demographics, type of road user, traumatic injuries, cause of death, and accident circumstances were extracted and supplemented by police reports. Comparative statistical analyses were performed for categorical and continuous variables. Results: Among 395 RTA autopsies, 23 (5.8%) involved children and 51 (12.9%) older adults. Most child victims were passengers (56.5%), whereas elderly fatalities occurred mainly among pedestrians (33.3%) and cyclists (25.5%), with statistically significant differences between age groups. Polytrauma was the leading cause of death in both categories, though isolated cranio-cerebral trauma was proportionally more frequent in children. Crash circumstances also showed age-related patterns, with children more involved in high-energy collisions and older adults more frequently struck as pedestrians. Survival intervals showed a similar distribution across groups. Conclusions: Child and elderly RTA fatalities in Romania share common determinants, primarily driver-related behaviors and insufficient safety measures, while also exhibiting distinct age-related vulnerabilities. Autopsy-based data highlights these patterns and can guide targeted interventions such as stricter law enforcement, public education, and infrastructure improvements. Full article

► Show Figures

Figure 1

17 pages, 1293 KB

Open AccessArticle

Effects of Using a 360-Degree Swaying Chair on Physical Workload During VDT Work

by Sayaka Noda, Toshihisa Doi and Kuniko Yamashita

Safety 2025, 11(4), 124; https://doi.org/10.3390/safety11040124 - 15 Dec 2025

Abstract

Prolonged visual display terminal (VDT) work leads to static muscular loading, increasing the risk of musculoskeletal disorders. Active chairs have been proposed to alleviate such issues; however, solutions like balance balls often induce discomfort due to excessive instability. To address this trade-off, a [...] Read more.

Prolonged visual display terminal (VDT) work leads to static muscular loading, increasing the risk of musculoskeletal disorders. Active chairs have been proposed to alleviate such issues; however, solutions like balance balls often induce discomfort due to excessive instability. To address this trade-off, a 360° swaying chair was developed, though its physiological effects during VDT work remain unclear. This study aimed to investigate the effects of a 360° swaying chair on users performing VDT tasks. Two experiments compared the swaying chair with a standard office chair (OC) under two sitting postures: a forward tilt with feet forward (AC2) and with feet back (AC3). Muscle activity, motion analysis, and subjective evaluations were conducted. The results showed that the AC3 posture (feet back) better maintained the spinal S-curve and reduced activity in the thoracic and lumbar erector spinae and rectus abdominis compared to the AC2 posture and the OC, although it may increase lower-body load. A slight forward tilt promoted activation of the internal oblique muscle. Subjective comfort was not inferior to that of the OC. These findings suggest that the 360° swaying chair, particularly in the AC3 posture, can reduce upper-body muscular and postural loads during VDT work without compromising comfort. However, these findings should be interpreted as preliminary, as they are based on a small and homogeneous sample and short-term VDT tasks. Full article

► Show Figures

Figure 1

20 pages, 920 KB

Open AccessArticle

Analytical Assessment of Pedestrian Crashes on Low-Speed Corridors

by Therezia Matongo and Deo Chimba

Safety 2025, 11(4), 123; https://doi.org/10.3390/safety11040123 - 9 Dec 2025

Abstract

This study presents a comprehensive statewide analysis of pedestrian-involved crashes recorded in Tennessee between 2002 and 2025. We evaluated the influence of roadway, traffic, environmental, and socioeconomic factors on pedestrian crash frequency and severity with substantial components focused on lighting impacts including dark [...] Read more.

This study presents a comprehensive statewide analysis of pedestrian-involved crashes recorded in Tennessee between 2002 and 2025. We evaluated the influence of roadway, traffic, environmental, and socioeconomic factors on pedestrian crash frequency and severity with substantial components focused on lighting impacts including dark and nighttime. A multi-method analytical framework was implemented, combining descriptive statistics, non-parametric tests, regression analysis, and advanced machine learning techniques including the Adaptive Neuro-Fuzzy Inference System (ANFIS) and the gradient boosting model (XGBoost). Results indicated that dark and nighttime conditions accounted for a disproportionate share of severe crashes—fatal and serious injuries under dark conditions reached over 40%, compared to less than 20% during daylight. The statistical tests revealed statistically significant differences in both total injuries and fatalities between low-speed (≤35 mph) and higher-speed (40–45 mph) corridors. The regression result identified AADT and the number of lanes as the strongest predictors of crash frequency, showing that greater traffic exposure and wider cross-sections substantially elevate pedestrian risk, while terrain and peak-hour traffic exhibited negative associations with severe injuries. The XGBoost model, consisting of 300 trees, achieved R² = 0.857, in which the SHAP analysis revealed that AADT, the roadway functional class, and the number of lanes are the most influential variables. The ANFIS model demonstrated that areas with higher population density and greater proportions of households without vehicles experience more pedestrian crashes. These findings collectively establish how pedestrian crash risks are correlated with traffic exposure, roadway geometry, lighting, and socioeconomic conditions, providing a strong analytical foundation for data-driven safety interventions and policy development. Full article

(This article belongs to the Special Issue Safety of Vulnerable Road Users at Night)

► Show Figures

Figure 1

17 pages, 368 KB

Open AccessArticle

A Biomechanical Analysis of Posture and Effort During Computer Activities: The Role of Furniture

by María Fernanda Trujillo-Guerrero, William Venegas-Toro, Danni De la Cruz-Guevara, Iván Zambrano-Orejuela, Alvaro Page-Del Pozo and Silvia Santos-Cuadros

Safety 2025, 11(4), 122; https://doi.org/10.3390/safety11040122 - 9 Dec 2025

Abstract

The ergonomic risks associated with posture in conventional office workstations have been extensively studied, but there is limited research available on these risks in the context of home-based work environments. Most available studies rely solely on questionnaire-based statistical analyses, leaving a gap in [...] Read more.

The ergonomic risks associated with posture in conventional office workstations have been extensively studied, but there is limited research available on these risks in the context of home-based work environments. Most available studies rely solely on questionnaire-based statistical analyses, leaving a gap in understanding the specific conditions of home-based work environments. This study focuses on evaluating the effects of workstation conditions on posture and muscular efforts across three anatomical segments: head-neck, trunk-upper trapezius, and arm-deltoid. The analysis is conducted by simulating workstation setups commonly associated with academic activities performed by students during the COVID-19 pandemic. The conditions examined in this study include inadequate desk height, the use of chairs without armrests, and the use of laptops. Eighteen volunteers, comprising nine women and nine men, participated in experiments conducted under scenarios designed using a

2^{k}

statistical approach. In all experiments, participants completed questionnaires, and text-writing activities were performed to evaluate the effects of these conditions. This research introduces a new non-invasive technique for ergonomic assessment that integrates photogrammetry and surface electromyography (sEMG) to simultaneously evaluate posture and muscular effort. The developed methodology allows precise, contactless analysis of ergonomic conditions and can be adapted for various professional and academic teleworking environments. Significant effects were observed in the posture (°) of the trunk and head, with both small and large effects identified at significance levels of p < 0.001 under the furniture conditions studied. In terms of EMG activity, moderate effects were observed at p < 0.01 levels between table height and upper trapezius activation, while small effects were detected at p < 0.05 levels between the use of chairs without armrests and neck. Similarly, small to moderate effects were observed in the arm-deltoid segment under the same furniture conditions. These findings reveal information about the posture and muscular effort patterns associated with the studied tasks, offering knowledge that can be referenced for similar tasks in other technical fields where telematics activities are performed. Full article

► Show Figures

Figure 1

31 pages, 1941 KB

Open AccessArticle

Boosting Traffic Crash Prediction Performance with Ensemble Techniques and Hyperparameter Tuning

by Naima Goubraim, Zouhair Elamrani Abou Elassad, Hajar Mousannif and Mohamed Ameksa

Safety 2025, 11(4), 121; https://doi.org/10.3390/safety11040121 - 9 Dec 2025

Abstract

Road traffic crashes are a major global challenge, resulting in significant loss of life, economic burden, and societal impact. This study seeks to enhance the precision of traffic accident prediction using advanced machine learning techniques. This study employs an ensemble learning approach combining [...] Read more.

Road traffic crashes are a major global challenge, resulting in significant loss of life, economic burden, and societal impact. This study seeks to enhance the precision of traffic accident prediction using advanced machine learning techniques. This study employs an ensemble learning approach combining the Random Forest, the Bagging Classifier (Bootstrap Aggregating), the Extreme Gradient Boosting (XGBoost) and the Light Gradient Boosting Machine (LightGBM) algorithms. To address class imbalance and feature relevance, we implement feature selection using the Extra Trees Classifier and oversampling using the Synthetic Minority Over-sampling Technique (SMOTE). Rigorous hyperparameter tuning is applied to optimize model performance. Our results show that the ensemble approach, coupled with hyperparameter optimization, significantly improves prediction accuracy. This research contributes to the development of more effective road safety strategies and can help to reduce the number of road accidents. Full article

(This article belongs to the Special Issue Road Traffic Risk Assessment: Control and Prevention of Collisions)

► Show Figures

Figure 1

28 pages, 4585 KB

Open AccessArticle

Uncertainty-Aware Adaptive Intrusion Detection Using Hybrid CNN-LSTM with cWGAN-GP Augmentation and Human-in-the-Loop Feedback

by Clinton Manuel de Nascimento and Jin Hou

Safety 2025, 11(4), 120; https://doi.org/10.3390/safety11040120 - 5 Dec 2025

Abstract

Intrusion detection systems (IDSs) must operate under severe class imbalance, evolving attack behavior, and the need for calibrated decisions that integrate smoothly with security operations. We propose a human-in-the-loop IDS that combines a convolutional neural network and a long short-term memory network (CNN–LSTM) [...] Read more.

Intrusion detection systems (IDSs) must operate under severe class imbalance, evolving attack behavior, and the need for calibrated decisions that integrate smoothly with security operations. We propose a human-in-the-loop IDS that combines a convolutional neural network and a long short-term memory network (CNN–LSTM) classifier with a variational autoencoder (VAE)-seeded conditional Wasserstein generative adversarial network with gradient penalty (cWGAN-GP) augmentation and entropy-based abstention. Minority classes are reinforced offline via conditional generative adversarial (GAN) sampling, whereas high-entropy predictions are escalated for analysts and are incorporated into a curated retraining set. On CIC-IDS2017, the resulting framework delivered well-calibrated binary performance (ACC = 98.0%, DR = 96.6%, precision = 92.1%, F1 = 94.3%; baseline ECE ≈ 0.04, Brier ≈ 0.11) and substantially improved minority recall (e.g., Infiltration from 0% to >80%, Web Attack–XSS +25 pp, and DoS Slowhttptest +15 pp, for an overall +11 pp macro-recall gain). The deployed model remained lightweight (~42 MB, <10 ms per batch; ≈32 k flows/s on RTX-3050 Ti), and only approximately 1% of the flows were routed for human review. Extensive evaluation, including ROC/PR sweeps, reliability diagrams, cross-domain tests on CIC-IoT2023, and FGSM/PGD adversarial stress, highlights both the strengths and remaining limitations, notably residual errors on rare web attacks and limited IoT transfer. Overall, the framework provides a practical, calibrated, and extensible machine learning (ML) tier for modern IDS deployment and motivates future research on domain alignment and adversarial defense. Full article

► Show Figures

Graphical abstract

25 pages, 1853 KB

Open AccessArticle

Testing of a Safety Leadership Model

by Jian Shen and Maureen Hassall

Safety 2025, 11(4), 119; https://doi.org/10.3390/safety11040119 - 1 Dec 2025

Abstract

Fatal and serious injury rates remain unacceptably high in the construction industry. Leadership plays a critical role in safety management and serious and fatal injury prevention. However, limited research has examined industry practitioners’ perceptions of leadership and how it influences safety outcomes, particularly [...] Read more.

Fatal and serious injury rates remain unacceptably high in the construction industry. Leadership plays a critical role in safety management and serious and fatal injury prevention. However, limited research has examined industry practitioners’ perceptions of leadership and how it influences safety outcomes, particularly in the prevention of serious and fatal injuries in the construction industry. Therefore, a theoretical model for capturing perceptions of safety leadership was developed from a systematic literature review. To ensure that the labels and language used in the model can be understood by industry practitioners, a Delphi study was conducted involving twelve experts. Over three iterative rounds, the model was refined to include five leadership styles, seventeen elements, and eighty-five descriptive statements spanning the range from laissez-faire to transformational leadership. The refined model provides a comprehensive framework for understanding safety leadership and serves as a foundation for future empirical testing with frontline construction workers. Full article

► Show Figures

Figure 1

29 pages, 2341 KB

Open AccessArticle

A Prototype Risk Assessment Dashboard for the Construction Industry: Getting Experts in the Loop Thanks to Machine Learning

by Mirza Muntasir Nishat, Antoine Rauzy and Nils O. E. Olsson

Safety 2025, 11(4), 118; https://doi.org/10.3390/safety11040118 - 1 Dec 2025

Abstract

Construction work is fundamentally hazardous. Traditional risk assessment tools (e.g., checklists and audits) are static in essence and hard to make evolve. In this paper, we demonstrate how to get experts dynamically in the loop thanks to machine learning. Namely, we discussed the [...] Read more.

Construction work is fundamentally hazardous. Traditional risk assessment tools (e.g., checklists and audits) are static in essence and hard to make evolve. In this paper, we demonstrate how to get experts dynamically in the loop thanks to machine learning. Namely, we discussed the design of a prototype risk assessment dashboard dedicated to fall accidents. The interactive graphical user interface allows professionals to generate construction scenarios and compare their evaluation of risks with that of the dashboard. The latter continuously learns from expert feedback. The proof-of-concept we present here shows that it is possible to capitalize on expert knowledge in a dynamic and user-friendly way. Thanks to its neural network architecture, not only does the dashboard learn from the experts, but professionals also learn from the dashboard. Full article

► Show Figures

Figure 1

19 pages, 602 KB

Open AccessArticle

Examining the Effects of Sight Distance, Road Conditions, and Weather on Intersection Crash Severity: A Random Parameters Logit Approach with Heterogeneity in Means and Variances

by Irfan Ullah, Ahmed Farid and Khaled Ksaibati

Safety 2025, 11(4), 117; https://doi.org/10.3390/safety11040117 - 27 Nov 2025

Abstract

Intersections represent critical crash locations on road networks necessitating targeted safety interventions. This study employs a random parameters ordered logit (RPOL) model with heterogeneity in means to analyze injury severity contributing factors across 9108 Wyoming intersection crashes that occurred from 2007 to 2017. [...] Read more.

Intersections represent critical crash locations on road networks necessitating targeted safety interventions. This study employs a random parameters ordered logit (RPOL) model with heterogeneity in means to analyze injury severity contributing factors across 9108 Wyoming intersection crashes that occurred from 2007 to 2017. The analysis reveals that crashes on principal and minor arterial intersections are consistently associated with higher risks of severe/fatal injuries, while urban intersections exhibit less severe consequences, likely due to lower speeds and enhanced infrastructure. Adverse weather conditions, particularly snowy and icy road surfaces, increase the likelihood of property-damage-only (PDO) outcomes while reducing severe/fatal injuries. Temporal trends show a decline in crash severity over time, coinciding with advances in vehicle safety and policy improvements. Key behavioral factors, including left turn maneuvers and driver’s age heterogeneity, influence crash outcomes, whereas intersection sight distance (ISD) had no significant effect on crash severity underscoring data limitations requiring advanced analysis methods. This study’s findings prioritize the reconsideration of arterial intersection design, urban safety enhancements, and behavior-focused countermeasures for intersection safety. Full article

► Show Figures

Figure 1

13 pages, 1917 KB

Open AccessArticle

Occupational Ergonomic Risks Among Women in Underground Coal Mining, South Africa

by Ouma S. Mokwena, Joyce Shirinde and Thabiso J. Morodi

Safety 2025, 11(4), 116; https://doi.org/10.3390/safety11040116 - 25 Nov 2025

Cited by 1

Abstract

Although women have participated in mining activities across the world for centuries, the industry continues to be perceived as predominantly male-oriented. This perception persists largely due to the male-dominated workforce and the physically demanding nature of mining operations. This paper examines the ergonomic [...] Read more.

Although women have participated in mining activities across the world for centuries, the industry continues to be perceived as predominantly male-oriented. This perception persists largely due to the male-dominated workforce and the physically demanding nature of mining operations. This paper examines the ergonomic impacts of mining machinery on female mineworkers. The study was conducted in three underground coal mining operations located in Mpumalanga, South Africa, using a quantitative research approach. To evaluate the ergonomic demands placed on women working underground, the researchers employed the Rapid Entire Body Assessment (REBA) in combination with direct observation techniques. The findings revealed that female mineworkers experience considerable challenges when performing tasks requiring significant physical strength and endurance. The observed female mineworker recorded a final REBA score of seven, indicating a medium-risk level. Ergonomic challenges in underground coal mining are further intensified for female mineworkers due to the absence of gender-specific considerations in equipment design, task allocation, and the overall working environment. Although the risk classification was moderate, the results underscore the need for further investigation and the timely implementation of corrective measures. Addressing these issues will require the integration of inclusive ergonomic principles that account for gender diversity within the mining workforce. Full article

► Show Figures

Figure 1

19 pages, 1268 KB

Open AccessArticle

Assessing the Potential Impact of Fugitive Methane Emissions on Offshore Platform Safety

by Stuart N. Riddick, Mercy Mbua, Catherine Laughery and Daniel J. Zimmerle

Safety 2025, 11(4), 115; https://doi.org/10.3390/safety11040115 - 24 Nov 2025

Abstract

One of the biggest risks to safety on offshore platform safety is the ignition of high-pressure natural gas streams. Currently, the size and number of fugitive emissions on offshore platforms is unknown and methods used to detect fugitives have significant shortcomings. To investigate [...] Read more.

One of the biggest risks to safety on offshore platform safety is the ignition of high-pressure natural gas streams. Currently, the size and number of fugitive emissions on offshore platforms is unknown and methods used to detect fugitives have significant shortcomings. To investigate the frequency, size, and potential impact of fugitives, a data collection exercise was conducted using incidents reported, leak survey data, and independent measurements. The size and number of fugitives on offshore facilities were simulated to investigate likely areas of safety concern. Incident reports indicate in 2021 there were 113 reports of gas leaks on 1119 offshore facilities, suggesting 0.02 fugitives per Type 1 facility (older, shallow-water platforms) and 0.31 fugitives per Type 2 facility (larger deeper-water facilities). Leak survey data report 12 fugitives per Type 1 facility (average emission 0.6 kg CH₄ h⁻¹ leak⁻¹) and 15 fugitives per Type 2 facility (average emission 1.5 kg CH₄ h⁻¹ leak⁻¹). Reconciliation of direct measurements with a bottom-up model suggests that the number of fugitive emissions generated from the leak report data is an underestimate for Type 1 platforms (44 fugitives facility⁻¹; average emission 0.6 kg CH₄ h⁻¹ leak⁻¹) and in general agreement for the Type 2 platforms (15 fugitives facility⁻¹; average emission 1.5 kg CH₄ h⁻¹ leak⁻¹). Analysis of the fugitive emission rates on an offshore platform suggests that gas will not collect to explosive concentration if any air movement is present (>0.36 mph); however, large volumes of air (~600 m³) near representative leaks on the working deck could become explosive in hour-long zero-wind conditions. We suggest that wearable technology could be employed to indicate gas build up, safety regulations amended to consider low-wind conditions and real-world experiments are conducted to test assumptions of air mixing on the working deck. Full article

► Show Figures

Figure 1

Journal Description

Safety

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Topical Collections

Further Information

Guidelines

MDPI Initiatives

Follow MDPI