Pub Date : 2025-11-07DOI: 10.1016/j.ssci.2025.107046
Matthew Mapa , Yuran Sun , Erica D. Kuligowski , Thomas J. Cova , Ruggiero Lovreglio , Xilei Zhao
Knowledge of household evacuation behavior in wildfires is critical to creating effective emergency plans. Understanding evacuation route choices is especially important for both evacuation management strategies and traffic simulations during emergencies. Existing studies of evacuation route choice rely solely on survey or interview data, which have inherent limitations in spatiotemporal resolution and memory bias. This study proposes a new methodology to analyze evacuation route choices and complement existing data sources by leveraging a GPS dataset. The proposed methodology includes a framework for systematically handling sparse GPS data and testing common assumptions used by existing evacuation simulations, i.e. whether evacuees take the shortest path, make a single trip to their destination, or are at home at the start of the evacuation. We applied this new method to a sample of 155 evacuees from the 2021 Marshall Fire in Boulder County, CO, which resulted in the evacuation of over 30,000 people. We found that the majority of evacuees approximated the shortest path (67.1%), made multiple trips while evacuating to their final destination (64.7%), and were home at the start of the evacuation (60%). The findings of this study can be used to inform better plans for future emergencies and enhance traffic simulations of evacuation behavior.
{"title":"Analyzing evacuation route choices during the 2021 Marshall Fire","authors":"Matthew Mapa , Yuran Sun , Erica D. Kuligowski , Thomas J. Cova , Ruggiero Lovreglio , Xilei Zhao","doi":"10.1016/j.ssci.2025.107046","DOIUrl":"10.1016/j.ssci.2025.107046","url":null,"abstract":"<div><div>Knowledge of household evacuation behavior in wildfires is critical to creating effective emergency plans. Understanding evacuation route choices is especially important for both evacuation management strategies and traffic simulations during emergencies. Existing studies of evacuation route choice rely solely on survey or interview data, which have inherent limitations in spatiotemporal resolution and memory bias. This study proposes a new methodology to analyze evacuation route choices and complement existing data sources by leveraging a GPS dataset. The proposed methodology includes a framework for systematically handling sparse GPS data and testing common assumptions used by existing evacuation simulations, i.e. whether evacuees take the shortest path, make a single trip to their destination, or are at home at the start of the evacuation. We applied this new method to a sample of 155 evacuees from the 2021 Marshall Fire in Boulder County, CO, which resulted in the evacuation of over 30,000 people. We found that the majority of evacuees approximated the shortest path (67.1%), made multiple trips while evacuating to their final destination (64.7%), and were home at the start of the evacuation (60%). The findings of this study can be used to inform better plans for future emergencies and enhance traffic simulations of evacuation behavior.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107046"},"PeriodicalIF":5.4,"publicationDate":"2025-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145475245","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-06DOI: 10.1016/j.ssci.2025.107033
Masaki Shibata , Daiel Pieper , Bernadette Matthews , Hannah Calverley , Jiyoung Kim , Soyeon Kim , Lucien Brown
Background
Australia recorded 150 coastal drownings with approximately 9,000 rescues in 2023/2024. Over the period 2013–2023, 30% of the drowning deaths recorded were of people born overseas. South Koreans, who make up the 5th largest group of international visitors, are found to be overrepresented in these figures. Multiple safety strategies aim to prevent these incidents; this study explores South Koreans’ understanding of Australian beach safety signs and identifies ways to improve signage.
Research Method.
This study utilised an online survey consisting of closed and open questions relating to demographics, intended behaviours at beaches and interpretation of signage. 163 university students studying in South Korea participated in this study. The statistical data were analysed, using descriptive analysis in SPSS, and the textual data were analysed using deductive code analysis in NVivo.
Result
Only about 14% understood red and yellow safety flags correctly and the majority did not understand the signage terms including rip currents, and the Korean translation, iallyu. Over 90% indicated they would not enter the water when seeing the dangerous current sign, but the presence of other people in the water encouraged about half of participants to go into the water. All red-coloured signs denoting no swimming were equally perceived as a higher hazard, while a yellow sign, despite the message, beach closed, and dangerous current were perceived as lower.
Implications.
This study proposes how signage terms in English could be revised for better translatability, while suggesting that red colours should be included on signage to indicate greater hazard.
{"title":"South Korean perceptions of Australian beach safety Signage: Insights and Evidence-Based recommendations for Improvement","authors":"Masaki Shibata , Daiel Pieper , Bernadette Matthews , Hannah Calverley , Jiyoung Kim , Soyeon Kim , Lucien Brown","doi":"10.1016/j.ssci.2025.107033","DOIUrl":"10.1016/j.ssci.2025.107033","url":null,"abstract":"<div><h3>Background</h3><div>Australia recorded 150 coastal drownings with approximately 9,000 rescues in 2023/2024. Over the period 2013–2023, 30% of the drowning deaths recorded were of people born overseas. South Koreans, who make up the 5th largest group of international visitors, are found to be overrepresented in these figures. Multiple safety strategies aim to prevent these incidents; this study explores South Koreans’ understanding of Australian beach safety signs and identifies ways to improve signage.</div><div>Research Method.</div><div>This study utilised an online survey consisting of closed and open questions relating to demographics, intended behaviours at beaches and interpretation of signage. 163 university students studying in South Korea participated in this study. The statistical data were analysed, using descriptive analysis in SPSS, and the textual data were analysed using deductive code analysis in NVivo.</div></div><div><h3>Result</h3><div>Only about 14% understood red and yellow safety flags correctly and the majority did not understand the signage terms including <em>rip currents,</em> and the Korean translation, <em>iallyu.</em> Over 90% indicated they would not enter the water when seeing the <em>dangerous current</em> sign, but the presence of other people in the water encouraged about half of participants to go into the water. All red-coloured signs denoting <em>no swimming</em> were equally perceived as a higher hazard, while a yellow sign, despite the message, <em>beach closed,</em> and <em>dangerous current</em> were perceived as lower<em>.</em></div><div>Implications.</div><div>This study proposes how signage terms in English could be revised for better translatability, while suggesting that red colours should be included on signage to indicate greater hazard.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107033"},"PeriodicalIF":5.4,"publicationDate":"2025-11-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145475242","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-05DOI: 10.1016/j.ssci.2025.107045
Hongcheng Lu , Sisi Sun , Xueer Zhang , Congcong Ge , Ran Ye , Jinghong Wang , Ahmed Mebarki , Jialin Wu , Yi Yang
Staircases are critical escape routes in emergency evacuations. However, in-depth research on the mechanism by which “invisibility” alters our mind and body remains insufficient. This study quantifies individuals’ physiological and psychological states through staircase evacuation experiments comparing normal and complete invisibility scenarios, and integrates multimodal physiological monitoring (ECG, BVP, EDA, TEMP) with the PAD emotion scale for data collection. Innovatively, based on dynamic segmentation analysis of cardiac cycles at landing platforms, Principal Component Analysis (PCA) was employed to integrate multiple physiological indicators, resulting in the development of a high-precision quantitative model (R2 > 0.98). From the key findings of the experimental study, it can be drawn that under completely invisible conditions, the correlation between physiological indicators and cardiac cycles was significantly enhanced (correlation coefficients generally > 0.95). It has also been observed that the PCA-derived composite indicator F effectively characterized the evolution of physiological stress. Psychologically, completely invisible conditions led to reduced pleasure (P), increased arousal (A), and decreased dominance (D), with emotions deteriorating significantly from “anxiety” to “fear”. In contrast, under normal conditions, emotions remained relatively stable, exhibiting a “dependent”. The study revealed the neurobiological pathway of “environmental stimuli-physiological responses-emotional experiences”, where physiological stress drove emotional changes, while variations in emotional characteristics reciprocally validated physiological upheavals. Together, they formed a complete evidence chain for pedestrian stress responses under environmental pressure, providing critical quantitative foundations for risk assessment and emergency management.
{"title":"Unveiling the invisible stress: A multimodal psychological and physiological analysis of staircase evacuation under visibility loss","authors":"Hongcheng Lu , Sisi Sun , Xueer Zhang , Congcong Ge , Ran Ye , Jinghong Wang , Ahmed Mebarki , Jialin Wu , Yi Yang","doi":"10.1016/j.ssci.2025.107045","DOIUrl":"10.1016/j.ssci.2025.107045","url":null,"abstract":"<div><div>Staircases are critical escape routes in emergency evacuations. However, in-depth research on the mechanism by which “invisibility” alters our mind and body remains insufficient. This study quantifies individuals’ physiological and psychological states through staircase evacuation experiments comparing normal and complete invisibility scenarios, and integrates multimodal physiological monitoring (ECG, BVP, EDA, TEMP) with the PAD emotion scale for data collection. Innovatively, based on dynamic segmentation analysis of cardiac cycles at landing platforms, Principal Component Analysis (PCA) was employed to integrate multiple physiological indicators, resulting in the development of a high-precision quantitative model (R<sup>2</sup> > 0.98). From the key findings of the experimental study, it can be drawn that under completely invisible conditions, the correlation between physiological indicators and cardiac cycles was significantly enhanced (correlation coefficients generally > 0.95). It has also been observed that the PCA-derived composite indicator F effectively characterized the evolution of physiological stress. Psychologically, completely invisible conditions led to reduced pleasure (P), increased arousal (A), and decreased dominance (D), with emotions deteriorating significantly from “anxiety” to “fear”. In contrast, under normal conditions, emotions remained relatively stable, exhibiting a “dependent”. The study revealed the neurobiological pathway of “environmental stimuli-physiological responses-emotional experiences”, where physiological stress drove emotional changes, while variations in emotional characteristics reciprocally validated physiological upheavals. Together, they formed a complete evidence chain for pedestrian stress responses under environmental pressure, providing critical quantitative foundations for risk assessment and emergency management.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107045"},"PeriodicalIF":5.4,"publicationDate":"2025-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145475240","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Context. Mobile equipment (ME), such as loader, that runs over pedestrian workers (PWs) is still a major cause of fatal work accidents. Proximity detection devices (PDDs) are an avenue to mitigate these risks, yet there is a lack of comprehensive studies to inform their practical implementation. Aim. The aim is to provide a structured analysis of the factors contributing to ME-PW collisions and to evaluate the potential impact of integrating PDDs into the causality chain of these accidents. Methodology. An exhaustive database was created through analysis of investigation reports on fatal ME-PW collisions in Quebec, Canada, from 2013 to 2023 (n = 34). Contributing factors were structured using fault tree analysis (FTA). Results. ME-PW collisions account for 8 % of all available fatal accident reports. Two accident types stand out: dump trucks backing up to deliver earthwork (8/34) and hydraulic shovels used in earthwork operations (8/34). Reversing collisions were the most frequent (21/34), as were accidents occurring during the starting phase of the ME (21/34). Discussion. In 88 % of cases, the operator was unaware of the PW’s presence just before the accident. This finding confirms the strong potential of PDD as an additional layer of risk reduction. The FTA provides a novel perspective on the potential benefits of PDDs, while also highlighting the challenges of their effective integration into real-world operations. For instance, (1) ensuring coverage of the actual danger zone according to the movement, (2) providing highly targeted warnings, and (3) accounting for subcontracting and environmental conditions that may affect PDD performance.
{"title":"Serious and fatal collisions between mobile equipment and pedestrian workers in Quebec: A critical analysis of contributing factors and the Role of proximity detection devices in risk reduction","authors":"Damien Burlet-Vienney , François Gauthier , Chantal Gauvin , Caroline Jolly , Aida Haghighi , Firdaous Sekkay , Alireza Saidi , Sabrina Jocelyn , Yuvin Chinniah","doi":"10.1016/j.ssci.2025.107038","DOIUrl":"10.1016/j.ssci.2025.107038","url":null,"abstract":"<div><div><strong>Context.</strong> Mobile equipment (ME), such as loader, that runs over pedestrian workers (PWs) is still a major cause of fatal work accidents. Proximity detection devices (PDDs) are an avenue to mitigate these risks, yet there is a lack of comprehensive studies to inform their practical implementation. Aim. The aim is to provide a structured analysis of the factors contributing to ME-PW collisions and to evaluate the potential impact of integrating PDDs into the causality chain of these accidents. <strong>Methodology.</strong> An exhaustive database was created through analysis of investigation reports on fatal ME-PW collisions in Quebec, Canada, from 2013 to 2023 (n = 34). Contributing factors were structured using fault tree analysis (FTA). <strong>Results.</strong> ME-PW collisions account for 8 % of all available fatal accident reports. Two accident types stand out: dump trucks backing up to deliver earthwork (8/34) and hydraulic shovels used in earthwork operations (8/34). Reversing collisions were the most frequent (21/34), as were accidents occurring during the starting phase of the ME (21/34). <strong>Discussion.</strong> In 88 % of cases, the operator was unaware of the PW’s presence just before the accident. This finding confirms the strong potential of PDD as an additional layer of risk reduction. The FTA provides a novel perspective on the potential benefits of PDDs, while also highlighting the challenges of their effective integration into real-world operations. For instance, (1) ensuring coverage of the actual danger zone according to the movement, (2) providing highly targeted warnings, and (3) accounting for subcontracting and environmental conditions that may affect PDD performance.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107038"},"PeriodicalIF":5.4,"publicationDate":"2025-11-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145475239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-03DOI: 10.1016/j.ssci.2025.107039
Junseo Lee , Sunhwa Park , Sehyeon Oh , Byungchol Ma
This study investigates whether state-of-the-art multimodal large language models (LLMs) can autonomously perform the entire Hazard and Operability Studies (HAZOP) process without human intervention. Four LLMs—GPT4o, GPT4o-mini, LLAMA, and Gemini—were utilized to generate automatic HAZOP worksheets spanning dozens of pages from an identical piping and instrumentation diagram (P&ID) using a standardized prompt. Their outputs were benchmarked against an expert-prepared refence worksheet and evaluated along two key aspects: (1) model performance, measured by similarity and computational cost and (2) HAZOP performance, measured by the validity of generated scenario and the diversity of safeguards. The results showed that all four LLMs achieved high similarity scores to the reference (F1 scores > 86 %). LLAMA was the most cost-efficient ($0.011 per worksheet), while Gemini generated the greatest number of scenarios (34.3 per worksheet) and safeguards (1.79 per deviation). This study presents a structured framework for evaluating LLMs in HAZOP and highlights their potential as assistive tools in the process safety field. However, key limitations were observed. The proportion of semantically valid scenarios remained low (0.19 to 0.37), and safeguards were heavily biased toward procedural measures, indicating limited diversity in risk-mitigation strategies. To enhance the reliability and practicability of LLM-based HAZOP studies, future research should focus on advanced prompt engineering, domain-specific fine-tuning, and improved reasoning capabilities.
{"title":"Can large language models automate the HAZOP process without human intervention?","authors":"Junseo Lee , Sunhwa Park , Sehyeon Oh , Byungchol Ma","doi":"10.1016/j.ssci.2025.107039","DOIUrl":"10.1016/j.ssci.2025.107039","url":null,"abstract":"<div><div>This study investigates whether state-of-the-art multimodal large language models (LLMs) can autonomously perform the entire Hazard and Operability Studies (HAZOP) process without human intervention. Four LLMs—GPT4o, GPT4o-mini, LLAMA, and Gemini—were utilized to generate automatic HAZOP worksheets spanning dozens of pages from an identical piping and instrumentation diagram (P&ID) using a standardized prompt. Their outputs were benchmarked against an expert-prepared refence worksheet and evaluated along two key aspects: (1) model performance, measured by similarity and computational cost and (2) HAZOP performance, measured by the validity of generated scenario and the diversity of safeguards. The results showed that all four LLMs achieved high similarity scores to the reference (F1 scores > 86 %). LLAMA was the most cost-efficient ($0.011 per worksheet), while Gemini generated the greatest number of scenarios (34.3 per worksheet) and safeguards (1.79 per deviation). This study presents a structured framework for evaluating LLMs in HAZOP and highlights their potential as assistive tools in the process safety field. However, key limitations were observed. The proportion of semantically valid scenarios remained low (0.19 to 0.37), and safeguards were heavily biased toward procedural measures, indicating limited diversity in risk-mitigation strategies. To enhance the reliability and practicability of LLM-based HAZOP studies, future research should focus on advanced prompt engineering, domain-specific fine-tuning, and improved reasoning capabilities.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107039"},"PeriodicalIF":5.4,"publicationDate":"2025-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145475352","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-03DOI: 10.1016/j.ssci.2025.107040
Che Khairil Izam Che Ibrahim , Patrick Manu , Sheila Belayutham , Clara Cheung , Mazlina Zaira Mohammad , Akilu Yunusa-Kaltungo , Shafienaz Ismail , Nor Syamimi Samsudin , Abhinesh Prabhakaran , Pablo Perez , Abdul-Majeed Mahamadu , Azman Hussain
Despite the growing recognition of Design for Safety (DfS) as a proactive approach to reducing construction-related accidents, its integration into practice remains inconsistent and fragmented. This study proposes a Tripartite Co-Creation Framework that aligns the strategic roles of academia, industry, and government (AIG) to institutionalize systematic DfS integration. Grounded in the Triple Helix model and structured through the Collective Impact (CI) framework, the study draws from a systematic literature review (SLR) of 96 peer-reviewed publications to identify six interrelated challenges impeding DfS adoption. These challenges are reframed into six strategic collaboration areas: (1) Cultivating a Safety-Oriented Mindset Through Education and Practice, (2) Financing Safety as a Shared Investment, (3) Regulatory Synchronization Through Multi-Stakeholder Governance, (4) Integrating Stakeholders for Proactive Design Collaboration, (5) Advancing Smart Design Through Digital Innovation, and (6) Enabling Adaptive Learning Through Research and Feedback. Each collaboration area is restructured into collaborative domains supported by five CI conditions; shared agenda, measurement systems, reinforcing activities, continuous communication, and backbone support. The resulting framework delineates strategic roles and mutual responsibilities across AIG actors, offering actionable mechanisms for embedding DfS into their respective domain. The framework sets the foundation for future empirical validation and offers a scalable model for mainstreaming DfS in diverse construction ecosystems. The study contributes to DfS scholarship by offering a scalable and theoretically informed framework for coordinated systemic implementation.
{"title":"Tripartite collaboration for design for safety in construction: A co-creation framework integrating academia, industry, and government","authors":"Che Khairil Izam Che Ibrahim , Patrick Manu , Sheila Belayutham , Clara Cheung , Mazlina Zaira Mohammad , Akilu Yunusa-Kaltungo , Shafienaz Ismail , Nor Syamimi Samsudin , Abhinesh Prabhakaran , Pablo Perez , Abdul-Majeed Mahamadu , Azman Hussain","doi":"10.1016/j.ssci.2025.107040","DOIUrl":"10.1016/j.ssci.2025.107040","url":null,"abstract":"<div><div>Despite the growing recognition of Design for Safety (DfS) as a proactive approach to reducing construction-related accidents, its integration into practice remains inconsistent and fragmented. This study proposes a Tripartite Co-Creation Framework that aligns the strategic roles of academia, industry, and government (AIG) to institutionalize systematic DfS integration. Grounded in the Triple Helix model and structured through the Collective Impact (CI) framework, the study draws from a systematic literature review (SLR) of 96 peer-reviewed publications to identify six interrelated challenges impeding DfS adoption. These challenges are reframed into six strategic collaboration areas: (1) Cultivating a Safety-Oriented Mindset Through Education and Practice, (2) Financing Safety as a Shared Investment, (3) Regulatory Synchronization Through Multi-Stakeholder Governance, (4) Integrating Stakeholders for Proactive Design Collaboration, (5) Advancing Smart Design Through Digital Innovation, and (6) Enabling Adaptive Learning Through Research and Feedback. Each collaboration area is restructured into collaborative domains supported by five CI conditions; shared agenda, measurement systems, reinforcing activities, continuous communication, and backbone support. The resulting framework delineates strategic roles and mutual responsibilities across AIG actors, offering actionable mechanisms for embedding DfS into their respective domain. The framework sets the foundation for future empirical validation and offers a scalable model for mainstreaming DfS in diverse construction ecosystems. The study contributes to DfS scholarship by offering a scalable and theoretically informed framework for coordinated systemic implementation.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107040"},"PeriodicalIF":5.4,"publicationDate":"2025-11-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145475241","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-11-01DOI: 10.1016/j.ssci.2025.107041
Carlos Albarrán Morillo , Huxiao Shi , John Suárez-Pérez , Micaela Demichela
Physical fatigue in repetitive production lines contributes to musculoskeletal disorders and absenteeism. This study investigates a pharmaceutical packaging environment in Colombia with 43 operators (42 female; 19–53 years) performing repetitive inspection and packing. Smartwatches captured pulse rate, electrodermal activity, skin temperature, and motion, complemented by demographic (age, experience) and occupational factors (task load, line, shift, timing). Principal Component Analysis (PCA) reduced dimensionality, and a fuzzy logic–based labeling method—adapted from prior controlled experiments—generated binary and four-class fatigue labels without mid-shift self-reports. These labeled datasets were used to train multiple machine-learning classifiers. Integrating contextual features with biometrics substantially improved performance: in binary classification, F1 increased from 0.8848 (biometrics only) to 0.9375; in four-level classification, F1 rose from 0.8232 to 0.8793. Motion-related metrics emerged as the most informative predictors. Critically, feature integration improved reliability: accuracy for intermediate states (Higher Non-Fatigue and Higher Fatigue) rose by ∼10 percentage points, while false negatives in the Pure Fatigue class were eliminated—3% of cases previously misclassified as Higher Non-Fatigue were instead correctly mapped within the fatigue spectrum. This shift strengthens the system’s effectiveness for real-time safety interventions. The novelty of this work lies in combining biometric and contextual modeling to reduce false negatives in critical fatigue states, providing a scalable, non-intrusive, and human-centered early-warning system. By aligning with Industry 5.0, this approach demonstrates how wearable and contextual data can jointly support proactive and trustworthy safety interventions while maintaining operational flow.
{"title":"Early detection of physical fatigue in industry using wearable sensors and contextual modeling","authors":"Carlos Albarrán Morillo , Huxiao Shi , John Suárez-Pérez , Micaela Demichela","doi":"10.1016/j.ssci.2025.107041","DOIUrl":"10.1016/j.ssci.2025.107041","url":null,"abstract":"<div><div>Physical fatigue in repetitive production lines contributes to musculoskeletal disorders and absenteeism. This study investigates a pharmaceutical packaging environment in Colombia with 43 operators (42 female; 19–53 years) performing repetitive inspection and packing. Smartwatches captured pulse rate, electrodermal activity, skin temperature, and motion, complemented by demographic (age, experience) and occupational factors (task load, line, shift, timing). Principal Component Analysis (PCA) reduced dimensionality, and a fuzzy logic–based labeling method—adapted from prior controlled experiments—generated binary and four-class fatigue labels without mid-shift self-reports. These labeled datasets were used to train multiple machine-learning classifiers. Integrating contextual features with biometrics substantially improved performance: in binary classification, F1 increased from 0.8848 (biometrics only) to 0.9375; in four-level classification, F1 rose from 0.8232 to 0.8793. Motion-related metrics emerged as the most informative predictors. Critically, feature integration improved reliability: accuracy for intermediate states (Higher Non-Fatigue and Higher Fatigue) rose by ∼10 percentage points, while false negatives in the Pure Fatigue class were eliminated—3% of cases previously misclassified as Higher Non-Fatigue were instead correctly mapped within the fatigue spectrum. This shift strengthens the system’s effectiveness for real-time safety interventions. The novelty of this work lies in combining biometric and contextual modeling to reduce false negatives in critical fatigue states, providing a scalable, non-intrusive, and human-centered early-warning system. By aligning with Industry 5.0, this approach demonstrates how wearable and contextual data can jointly support proactive and trustworthy safety interventions while maintaining operational flow.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107041"},"PeriodicalIF":5.4,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145425816","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-30DOI: 10.1016/j.ssci.2025.107037
Yiqin Yu, Wenqi Li, Yao Wang, Dongping Fang
The significance of safety culture in construction is well established, yet its definition and measurement remain contested. Existing approaches often focus on organizational or behavioral interventions, overlooking how national policy texts function as cultural artifacts that encode institutional expectations. This study addresses this gap by examining how safety culture is constructed in regulatory texts using Latent Dirichlet Allocation and a theory-informed classification framework. Analysis of 45 construction safety policies from nine countries and three international organizations reveals three cultural pathways (values, cognition, and behavior), which are embedded in policy texts. These pathways are hierarchically structured and vary across jurisdictions, reflecting distinct governance styles and regulatory logics. Challenging the notion that safety culture is vague or externally imposed, the study demonstrates that policy texts serve as structured vehicles for encoding cultural meaning. It offers a scalable framework for extracting structured cultural meaning from regulatory documents and provides comparative insights that contribute to safety theory, policy analysis, and more context-sensitive regulatory design.
{"title":"Decoding construction safety culture in policy texts: A theory-guided hierarchical topic modeling research","authors":"Yiqin Yu, Wenqi Li, Yao Wang, Dongping Fang","doi":"10.1016/j.ssci.2025.107037","DOIUrl":"10.1016/j.ssci.2025.107037","url":null,"abstract":"<div><div>The significance of safety culture in construction is well established, yet its definition and measurement remain contested. Existing approaches often focus on organizational or behavioral interventions, overlooking how national policy texts function as cultural artifacts that encode institutional expectations. This study addresses this gap by examining how safety culture is constructed in regulatory texts using Latent Dirichlet Allocation and a theory-informed classification framework. Analysis of 45 construction safety policies from nine countries and three international organizations reveals three cultural pathways (values, cognition, and behavior), which are embedded in policy texts. These pathways are hierarchically structured and vary across jurisdictions, reflecting distinct governance styles and regulatory logics. Challenging the notion that safety culture is vague or externally imposed, the study demonstrates that policy texts serve as structured vehicles for encoding cultural meaning. It offers a scalable framework for extracting structured cultural meaning from regulatory documents and provides comparative insights that contribute to safety theory, policy analysis, and more context-sensitive regulatory design.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107037"},"PeriodicalIF":5.4,"publicationDate":"2025-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145425815","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
During the last decades, a significant change has taken place in the demographic characteristics of the workforce: a growing percentage of employees is ageing, and organizations are beginning to feel the impact of an Ageing Workforce (AW) on production system performance. Yet this aspect can play a significant role in effective Occupational Safety and Health (OSH) decision-making, it is often underestimated or overlooked. Hence, an exploratory questionnaire-based study—grounded in the existing literature—maps common OSH interventions in an AW context and assesses their impact on OSH performance and productivity. Additionally, the study investigates key drivers and barriers—both ageing-specific and more general—that shape, facilitate, or limit intervention outcomes. Findings reveal that targeted communication and training interventions, along with ergonomic and technical tools, most strongly enhance OSH performance and productivity for AWs. Key drivers include active worker involvement and management commitment, whereas barriers such as poor safety culture, limited risk awareness, and legislative complexity hinder success. Overall, while economic incentives and policy compliance matter, they are not sufficient to drive sustainable improvements in OSH performance. Instead, internal organizational dynamics—especially cultural and relational factors that emphasize stability, recognition, and meaningful participation—are the ones that most influence the successful deployment of effective, inclusive interventions in an AW context.
{"title":"An exploratory study of workplace Occupational Safety and Health Interventions and their impact on productivity and performance in an Ageing Workforce","authors":"Gaia Vitrano , Simone Arena , Cleo Varianou-Mikellidou , Guido J.L. Micheli , Fabio Sgarbossa , Georgios Boustras","doi":"10.1016/j.ssci.2025.107031","DOIUrl":"10.1016/j.ssci.2025.107031","url":null,"abstract":"<div><div>During the last decades, a significant change has taken place in the demographic characteristics of the workforce: a growing percentage of employees is ageing, and organizations are beginning to feel the impact of an Ageing Workforce (AW) on production system performance. Yet this aspect can play a significant role in effective Occupational Safety and Health (OSH) decision-making, it is often underestimated or overlooked. Hence, an exploratory questionnaire-based study—grounded in the existing literature—maps common OSH interventions in an AW context and assesses their impact on OSH performance and productivity. Additionally, the study investigates key drivers and barriers—both ageing-specific and more general—that shape, facilitate, or limit intervention outcomes. Findings reveal that targeted communication and training interventions, along with ergonomic and technical tools, most strongly enhance OSH performance and productivity for AWs. Key drivers include active worker involvement and management commitment, whereas barriers such as poor safety culture, limited risk awareness, and legislative complexity hinder success. Overall, while economic incentives and policy compliance matter, they are not sufficient to drive sustainable improvements in OSH performance. Instead, internal organizational dynamics—especially cultural and relational factors that emphasize stability, recognition, and meaningful participation—are the ones that most influence the successful deployment of effective, inclusive interventions in an AW context.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107031"},"PeriodicalIF":5.4,"publicationDate":"2025-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145425814","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-10-29DOI: 10.1016/j.ssci.2025.107042
Soyeon Kim, Simeon Calvert, Marjan Hagenzieker
As Advanced Driver Assistance Systems (ADAS) become integrated into vehicles, driver education is important to support the safe and effective use of these technologies. However, structured ADAS educational programs for drivers have not been extensively studied. Moreover, the perspective of driving instructors, key stakeholders in the training process, has been overlooked. To address this gap, this study explores the perspectives of professional driving instructors who have delivered structured ADAS driver training at driving academies across four European countries. Through semi-structured interviews with fourteen instructors, this study examines the impact of the training, training design, implementation challenges, demographic considerations, and institutional roles. Instructors reported that ADAS driver training enhances driver confidence and promotes the appropriate use of the system, particularly by reducing overreliance on automation. They also emphasised the importance of a phased training model, combining theoretical instruction, controlled on-track practice, and on-road driving. In addition, Instructors highlighted the need for tailored approaches for older drivers and for introducing ADAS training after novice drivers have acquired basic driving skills. This study suggests the need for standardised ADAS training and cross-sector collaboration among leasing companies, car dealerships, and regulatory bodies to ensure broad accessibility and effective learning. The findings contribute to developing scalable, inclusive, and safety-oriented frameworks for driver education in emerging vehicle technologies.
{"title":"Exploring ADAS driver training in driving academies: Perspectives from driving instructors","authors":"Soyeon Kim, Simeon Calvert, Marjan Hagenzieker","doi":"10.1016/j.ssci.2025.107042","DOIUrl":"10.1016/j.ssci.2025.107042","url":null,"abstract":"<div><div>As Advanced Driver Assistance Systems (ADAS) become integrated into vehicles, driver education is important to support the safe and effective use of these technologies. However, structured ADAS educational programs for drivers have not been extensively studied. Moreover, the perspective of driving instructors, key stakeholders in the training process, has been overlooked. To address this gap, this study explores the perspectives of professional driving instructors who have delivered structured ADAS driver training at driving academies across four European countries. Through semi-structured interviews with fourteen instructors, this study examines the impact of the training, training design, implementation challenges, demographic considerations, and institutional roles. Instructors reported that ADAS driver training enhances driver confidence and promotes the appropriate use of the system, particularly by reducing overreliance on automation. They also emphasised the importance of a phased training model, combining theoretical instruction, controlled on-track practice, and on-road driving. In addition, Instructors highlighted the need for tailored approaches for older drivers and for introducing ADAS training after novice drivers have acquired basic driving skills. This study suggests the need for standardised ADAS training and cross-sector collaboration among leasing companies, car dealerships, and regulatory bodies to ensure broad accessibility and effective learning. The findings contribute to developing scalable, inclusive, and safety-oriented frameworks for driver education in emerging vehicle technologies.</div></div>","PeriodicalId":21375,"journal":{"name":"Safety Science","volume":"194 ","pages":"Article 107042"},"PeriodicalIF":5.4,"publicationDate":"2025-10-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145425785","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号