Hasil untuk "Industrial safety. Industrial accident prevention"

Ahmed Basager, Abdullah Alrabghi

Musculoskeletal disorders remain a persistent concern in automotive repair, yet empirical evidence on task-specific ergonomic risks in Middle Eastern contexts is limited. This study provides a detailed ergonomic risk profile of auto body specialists in Jeddah, Saudi Arabia, using a mixed-method approach that integrates the Rapid Upper Limb Assessment (RULA), Rapid Entire Body Assessment (REBA), and a validated Nordic Musculoskeletal Questionnaire. Twenty-five specialists across diverse tasks including installation, weighing, painting, cutting, and lifting were systematically evaluated to identify both postural and self-reported risk patterns. Results showed a high prevalence of discomfort in the lower back (64%), shoulders (52%), and wrists (48%). Ergonomic assessment revealed that the evaluated tasks were predominantly classified as moderate-to-high-risk, with RULA scores ranging from 6 to 7 and REBA scores ranging from 8 to 11. Beyond confirming the physical strain inherent to auto body work, the study highlights contextual factors such as prolonged static postures, limited use of mechanical aids, and constrained workshop layouts that exacerbate ergonomic risks. Importantly, the findings inform multi-level recommendations ranging from workshop practices to industry standards and policy considerations ensuring that interventions are both practical and scalable. By situating locally grounded results within the broader discourse on musculoskeletal risk prevention, the study offers region-specific evidence while providing globally relevant lessons for labor-intensive industries.

Jiyoung Park, D. Kang

The integration of Artificial Intelligence (AI) and smart technologies into safety management is a pivotal aspect of the Fourth Industrial Revolution or Industry 4.0. This study conducts a systematic literature review to identify and analyze how AI and smart technologies enhance safety management across various sectors within the Safety 4.0 paradigm. Focusing on peer-reviewed journal articles that explicitly mention “Smart”, “AI”, or “Artificial Intelligence” in their titles, the research examines key safety management factors, such as accident prevention, risk management, real-time monitoring, and ethical implementation, across sectors, including construction, industrial safety, disaster and public safety, transport and logistics, energy and power, health, smart home and living, and other diverse industries. AI-driven solutions, such as predictive analytics, machine learning algorithms, IoT sensor integration, and digital twin models, are shown to proactively identify and mitigate potential hazards, optimize energy consumption, and enhance operational efficiency. For instance, in the energy and power sector, intelligent gas meters and automated fire suppression systems manage gas-related risks effectively, while in the health sector, AI-powered health monitoring devices and mental health support applications improve patient and worker safety. The analysis reveals a significant trend towards shifting from reactive to proactive safety management, facilitated by the convergence of AI with IoT and Big Data analytics. Additionally, ethical considerations and data privacy emerge as critical challenges in the adoption of AI technologies. The study highlights the transformative role of AI in enhancing safety protocols, reducing accident rates, and improving overall safety outcomes across industries. It underscores the need for standardized protocols, robust AI governance frameworks, and interdisciplinary research to address existing challenges and maximize the benefits of AI in safety management. Future research directions include developing explainable AI models, enhancing human–AI collaboration, and fostering global standardization to ensure the responsible and effective implementation of AI-driven safety solutions.

O. Oyeyipo, S. Adekunle, Andrew Ebekozien et al.

The construction industry is regarded as one of the most dangerous industries to work. This is due to its high accident and fatality rate. The birth of the fourth industrial revolution provided many technologies to enhance and foresee safety by preventing accidents. Examples of such technologies are augmented reality and virtual reality. The study assessed the Nigerian construction industry and identified the impediments to the adoption of augmented reality (AR) and virtual reality (VR) to enhance safety on construction project sites. Data was retrieved through a well-structured questionnaire administered randomly to construction industry professionals in Nigeria. Data collected was analysed and, the impediments clustered. The barriers were identified and clustered under - “Policy and the nature of the industry” and “Knowledge related factors”. The study identified and classified the impediments to the adoption of AR and VR for safety enhancement and accident prevention in the Nigerian construction industry. Also, the study proposed recommendations to achieve a safer construction industry.

Kawtar Benderouach, Idriss Bennis, Abdelouahad Bellat et al.

This paper aims to develop a machine learning based predictive model for detecting fatal and non-fatal industrial accidents in the energy sector. Four machine learning models such as Support Vector Machine (SVM), Logistic Regression (LR), Naïve Bayes (NB) and Random Forest (RF) were evaluated to identify the most suitable model. In addition, three feature extraction models—Bag of Words (BoW), TF-IDF and Word2Vec—were applied with the machine learning models to obtain a powerful model for fatal and non-fatal accident classification. The most popular metrics—accuracy, precision, F1-score and recall—were employed to assess the models. Testing confirmed that the achievement level amounted to 97% success. The investigation establishes the capability of machine learning to support energy sector safety management operations. The research results can support accident prevention through risk factor identification and safety hazard awareness improvement while enabling quantitative fatal and non-fatal accident predictions for better safety management system implementation.

Tita Alissa Bach, Linn Pedersen, Maria Kinck Borén† et al.

As cyber threats increasingly exploit human behaviour, technical controls alone cannot ensure organisational cybersecurity (CS). Strengthening cybersecurity culture (CSC) is vital in safety-critical industries, yet empirical research in real-world industrial setttings is scarce. This paper addresses this gap through a pilot mixed-methods CSC assessment in a global safety-critical organisation. We examined employees' CS knowledge, attitudes, behaviours, and organisational factors shaping them. A survey and semi-structured interviews were conducted at a global organisation in safety-critical industries, across two countries chosen for contrasting phishing simulation performance: Country 1 stronger, Country 2 weaker. In Country 1, 258 employees were invited (67%), in Country 2, 113 were invited (30%). Interviews included 20 and 10 participants respectively. Overall CSC profiles were similar but revealed distinct challenges. Both showed strong phishing awareness and prioritised CS, yet most viewed phishing as the main risk and lacked clarity on handling other incidents. Line managers were default contacts, but follow-up on reported concerns was unclear. Participants emphasized aligning CS expectations with job relevance and workflows. Key contributors to differences emerged: Country 1 had external employees with limited access to CS training and policies, highlighting monitoring gaps. In Country 2, low survey response stemmed from a "no-link in email" policy. While this policy may have boosted phishing performance, it also underscored inconsistencies in CS practices. Findings show that resilient CSC requires leadership involvement, targeted communication, tailored measures, policy-practice alignment, and regular assessments. Embedding these into strategy complements technical defences and strengthens sustainable CS in safety-critical settings.

Igor Shovkoshytnyi, Andrii Marchenko, Liudmyla Minenko

Під час поширення електромагнітних хвиль у реальних умовах спостерігається ефект їх затухання, що пов’язаний з поглинанням або розсіюванням хвиль у тропосфері, а також дифракційними втратами енергії. Це значно впливає на енергетичні параметри радіосигналів, визначення яких вважається особливо складним на межі зони радіовидимості. Метою статті є розроблення методу визначення енергетичних параметрів радіозавад (радіосигналів) на межі зони радіовидимості з використанням методу лінійної інтерполяції для забезпечення оперативних розрахунків з оцінювання електромагнітної обстановки. Методи дослідження. Під час написання статті застосовано методи системного аналізу процесів поширення електромагнітних хвиль та синтезу методу визначення енергетичних параметрів радіосигналів. Зазначений методологічний підхід дав змогу отримати вираз для визначення множника ослаблення електромагнітного поля, який враховує поширення електромагнітних хвиль у приземних шарах тропосфери з відповідними втратами, зокрема, завдяки різній поляризації сигналів і приймальних антен. Отримані результати дослідження. Запропоновано використовувати метод лінійної інтерполяції для визначення енергетичних параметрів радіосигналів на межі зони радіовидимості, на які впиває діапазон дальностей, що залежать від дальності прямої видимості. Встановлено, що енергетичні параметри радіозавад можуть оцінюватися через напруженість електромагнітного поля в заданій точці простору або густину потоку потужності (вектор Умова-Пойнтинга) радіосигналів у місці розташування приймальної антени. Отримано вираз для апроксимованого значення множника ослаблення електромагнітного поля на довільній дальності в межах «зони напівтіні» та розроблено порядок визначення енергетичних параметрів радіосигналів на межі зони радіовидимості. Елементом наукової новизни є те, що запропонований метод доцільно застосовувати для вирішення прикладних завдань побудови ліній радіозв'язку, а також для прогнозування можливої завадової обстановки в районі бойових дій. Теоретична та практична значущість статті. Використання методу може забезпечити підвищення оперативності визначення енергетичних параметрів радіозавад на межі зони радіовидимості завдяки зменшенню кількості обчислювальних процедур. Такий підхід може бути доцільним під час прогнозування можливої завадової обстановки в районі бойових дій.

Ana Marija Dunaj

Maria H. Pedrosa, Ana K. Salazar, Carla Cardoso et al.

Safety culture is crucial for organisations aiming to enhance safety performance and is challenging in traditional sectors. This study explored the effects of a new near-miss management system (NMS) on safety culture in traditional manufacturing companies. The data collection followed a mixed-methods design: quantitative data were collected through pre- and post-implementation surveys, and qualitative data were derived from focus group discussions (FGDs) developed a year after NMS implementation in two footwear companies. After that period, it is possible to infer that the NMS led to changes in the safety culture. OHS management commitment, accident and near-miss investigation, and OHS meetings and training dimensions were impacted. Employees perceived the NMS introduction as a signal of management′s commitment and the possibility of accessing training and participating in near-miss and accident investigations. In organisations that rely on intensive manual labour, ongoing training is essential to ensure that safety measures are effective and that the organisation′s safety management system (NMS) is sustainable. Although limited by the small sample size and sector focus, the results show that even a simplified NMS procedure adapted to the company and adequate training provided to the workforce significantly impacts the company′s safety culture and workers′ safety behaviour.

Heui Jung Park

This study analyzed the impact of the Occupational Safety and Health Act and the Serious Accident Punishment Act on industrial accident prevention and the criminal penalty system, focusing on judicial precedents to clearly identify practical changes in workplace safety management and corporate accountability. The objective was to comparatively analyze differences and complementarities between the two laws to establish a more effective legal framework for industrial accident prevention. The scope of the study included a theoretical review of concepts and prior research related to industrial and serious accidents, along with an analysis of the legal principles and applicability of the Occupational Safety and Health Act (2020-2024) and the Serious Accident Punishment Act (2022-2024). The research methodology employed literature review, case analysis, and comparative legal studies. Legal standards and sentencing trends applied by courts were derived using literature, reports, court rulings, and data from the Korean Law Information Center and judiciary sources. Furthermore, a comparative legal approach was adopted to thoroughly examine the complementarity between the two laws. The findings revealed that the Occupational Safety and Health Act had limitations in clearly defining responsible parties and the scope of its application, thereby restricting effective accountability and punishment. Conversely, since the implementation of the Serious Accident Punishment Act, courts have interpreted criminal elements (such as commission or omission of duties, causality, intent, and negligence) more rigorously, reinforcing corporate accountability and safety management. This demonstrated the necessity for complementary application of both laws to enhance compliance with safety measures in workplaces. In conclusion, the study clarified the characteristics of both laws and underscored the importance of establishing a complementary and effective legal framework. This framework addresses the shortcomings of existing legal applications and enhances industrial accident prevention and worker protection. Recommendations highlight the need for continuous improvement of practical legal applications and emphasize the importance of social discourse and collaborative institutional improvements among diverse stakeholders. The limitation noted in this research was the relatively short period of case law analyzed, suggesting a need for additional case accumulation and subsequent longitudinal analyses for a broader perspective on evolving legal and practical trends.

M.B. Aripkhodzhaeva, G.M. Gulomova, D. M. Rakhmatova et al.

The article examines the main causes of technological emergencies at industrial enterprises and proposes effective methods for their prevention. The analysis showed that most accidents are associated with human factors, equipment wear, and violations of safety protocols. Special attention is given to the influence of space-related factors, such as cosmic dust, radiation, and meteorite impacts, which can indirectly or directly affect the operation of high-tech equipment and automation systems. The author proposes the implementation of an early warning and accident prevention system based on continuous monitoring of equipment condition, consideration of external space factors, and the promotion of a safety culture in production.

Ho Doung Cho, J. Lee, Byeong Cheol Jeong

This study empirically investigates the effectiveness of a safety leadership coaching program targeting managers and executives in the manufacturing sector. In the context of increasing demands for serious accident prevention and the establishment of a robust safety culture in industrial settings, the development of managerial safety leadership capabilities has become essential. To address this need, a six-week coaching program based on the GROW model was developed and implemented. The program aimed to enhance field-oriented leadership competencies and promote safety behavior among employees. A total of 158 managers and 830 employees from manufacturing sites participated in the study. Using a pre-post design, changes in leadership behaviors and employee safety actions were measured through validated survey instruments. Data were analyzed using reliability testing, exploratory and confirmatory factor analyses, and structural equation modeling (SEM) to test the hypothesized relationships. The results revealed that participation in the coaching program significantly improved the managers’ field-oriented safety leadership, which in turn positively influenced employees’ proactive and participative safety behaviors. Moreover, the analysis confirmed that changes in managerial leadership served as a mediating factor between the coaching intervention and employees’ safety behavior, thereby contributing to the enhancement of the overall organizational safety culture. This study provides empirical evidence that coaching-based leadership development for managers can effectively drive actual changes in employee safety behavior. It further emphasizes the importance of moving beyond short-term, lecture-based training toward sustainable, coaching-driven interventions to advance safety leadership and foster a mature safety culture in industrial organizations.

Inaki Azarate, W. Buttner, H. Barthélemy et al.

A workshop outputs shaping international activities on hydrogen safety- covering Industrial and National Programmes; Applications; Storage; Accident Physics – Gas Phase; Accident Physics – Liquid/ Cryogenic Behaviour; Materials; Mitigation, Sensors, Hazard Prevention and Risk Reduction; Integrated Tools for Hazard and Risk Assessment; General Aspects of Safety. This report gives an overview of each topic made by the session chairperson. It also gives further analysis of the totality of the evidence presented.

Lele Feng, Yifan Gu, Jiabao Pang et al.

The hydrogen production from natural gas has advantages of low investment, low carbon emission, and high hydrogen production rate. This paper briefly describes the technical overview of hydrogen production from natural gas reforming and identifies its risk factors. According to the dangerous characteristics of high reaction temperature, easy leakage of reaction medium, flammability, and explosion in the process, the intrinsic safety of the process is discussed in combination with relevant research and industrial experience. The safety requirements of key equipment and materials are introduced in detail, followed by the optimization methods of process safety that can be taken in the engineering process. Besides, the accident prevention measures for emergency shutdown and fire explosion are summarized. Finally, the future research demands are put forward from the perspective of research and development, which is instructive for the safe hydrogen production from natural gas in the future.

Tieyi Yan, Jiaxin Wu, Munish Kumar et al.

The rapid process of urbanization and industrial development has raised significant concerns regarding the presence and management of hazardous substances. However, conventional methods employed for identifying hazardous substances and monitoring urban safety often suffer from low efficiency and accuracy. This paper proposes a novel approach that combines deep learning and genetic algorithms, which utilizes the Bidirectional Long Short-Term Memory model to capture temporal features in hazardous substance data and introduces the Attention Mechanism for weighted processing of crucial information, thereby improving recognition capability. Genetic Algorithms are employed to optimize the performance and generalization capacity of the deep learning model. Experimental validation demonstrates that the proposed approach achieves higher accuracy and faster processing speed, effectively enhancing urban safety monitoring. This research holds practical implications for urban safety management and accident prevention, offering an innovative solution to guarantee urban safety.

Zuoyong Li, Qinghua Lin, Haoyi Fan et al.

Semi-supervised learning suffers from the imbalance of labeled and unlabeled training data in the video surveillance scenario. In this paper, we propose a new semi-supervised learning method called SIAVC for industrial accident video classification. Specifically, we design a video augmentation module called the Super Augmentation Block (SAB). SAB adds Gaussian noise and randomly masks video frames according to historical loss on the unlabeled data for model optimization. Then, we propose a Video Cross-set Augmentation Module (VCAM) to generate diverse pseudo-label samples from the high-confidence unlabeled samples, which alleviates the mismatch of sampling experience and provides high-quality training data. Additionally, we construct a new industrial accident surveillance video dataset with frame-level annotation, namely ECA9, to evaluate our proposed method. Compared with the state-of-the-art semi-supervised learning based methods, SIAVC demonstrates outstanding video classification performance, achieving 88.76\% and 89.13\% accuracy on ECA9 and Fire Detection datasets, respectively. The source code and the constructed dataset ECA9 will be released in \url{https://github.com/AlchemyEmperor/SIAVC}.

Bjarke Madsen, Ramoni Adeogun

Recently, 6G in-X subnetworks have been proposed as low-power short-range radio cells to support localized extreme wireless connectivity inside entities such as industrial robots, vehicles, and the human body. Deployment of in-X subnetworks within these entities may result in rapid changes in interference levels and thus, varying link quality. This paper investigates distributed dynamic channel allocation to mitigate inter-subnetwork interference in dense in-factory deployments of 6G in-X subnetworks. This paper introduces two new techniques, Federated Multi-Agent Double Deep Q-Network (F-MADDQN) and Federated Multi-Agent Deep Proximal Policy Optimization (F-MADPPO), for channel allocation in 6G in-X subnetworks. These techniques are based on a client-to-server horizontal federated reinforcement learning framework. The methods require sharing only local model weights with a centralized gNB for federated aggregation thereby preserving local data privacy and security. Simulations were conducted using a practical indoor factory environment proposed by 5G-ACIA and 3GPP models for in-factory environments. The results showed that the proposed methods achieved slightly better performance than baseline schemes with significantly reduced signaling overhead compared to the baseline solutions. The schemes also showed better robustness and generalization ability to changes in deployment densities and propagation parameters.

Li Dong, Yubo Peng, Feibo Jiang et al.

In fire surveillance, Industrial Internet of Things (IIoT) devices require transmitting large monitoring data frequently, which leads to huge consumption of spectrum resources. Hence, we propose an Industrial Edge Semantic Network (IESN) to allow IIoT devices to send warnings through Semantic communication (SC). Thus, we should consider (1) Data privacy and security. (2) SC model adaptation for heterogeneous devices. (3) Explainability of semantics. Therefore, first, we present an eXplainable Semantic Federated Learning (XSFL) to train the SC model, thus ensuring data privacy and security. Then, we present an Adaptive Client Training (ACT) strategy to provide a specific SC model for each device according to its Fisher information matrix, thus overcoming the heterogeneity. Next, an Explainable SC (ESC) mechanism is designed, which introduces a leakyReLU-based activation mapping to explain the relationship between the extracted semantics and monitoring data. Finally, simulation results demonstrate the effectiveness of XSFL.

Daniel Barros, Paula Fraga-Lamas, Tiago M. Fernandez-Carames et al.

The Industry 5.0 paradigm focuses on industrial operator well-being and sustainable manufacturing practices, where humans play a central role, not only during the repetitive and collaborative tasks of the manufacturing process, but also in the management of the factory floor assets. Human factors, such as ergonomics, safety, and well-being, push the human-centric smart factory to efficiently adopt novel technologies while minimizing environmental and social impact. As operations at the factory floor increasingly rely on collaborative robots (CoBots) and flexible manufacturing systems, there is a growing demand for redundant safety mechanisms (i.e., automatic human detection in the proximity of machinery that is under operation). Fostering enhanced process safety for human proximity detection allows for the protection against possible incidents or accidents with the deployed industrial devices and machinery. This paper introduces the design and implementation of a cost-effective thermal imaging Safety Sensor that can be used in the scope of Industry 5.0 to trigger distinct safe mode states in manufacturing processes that rely on collaborative robotics. The proposed Safety Sensor uses a hybrid detection approach and has been evaluated under controlled environmental conditions. The obtained results show a 97% accuracy at low computational cost when using the developed hybrid method to detect the presence of humans in thermal images.

Xiang Li, Lan Zhao, Junhao Ren et al.

Effective information gathering and knowledge codification are pivotal for developing recommendation systems that promote circular economy practices. One promising approach involves the creation of a centralized knowledge repository cataloguing historical waste-to-resource transactions, which subsequently enables the generation of recommendations based on past successes. However, a significant barrier to constructing such a knowledge repository lies in the absence of a universally standardized framework for representing business activities across disparate geographical regions. To address this challenge, this paper leverages Large Language Models (LLMs) to classify textual data describing economic activities into the International Standard Industrial Classification (ISIC), a globally recognized economic activity classification framework. This approach enables any economic activity descriptions provided by businesses worldwide to be categorized into the unified ISIC standard, facilitating the creation of a centralized knowledge repository. Our approach achieves a 95% accuracy rate on a 182-label test dataset with fine-tuned GPT-2 model. This research contributes to the global endeavour of fostering sustainable circular economy practices by providing a standardized foundation for knowledge codification and recommendation systems deployable across regions.

Kailai Sun, Tianxiang Lan, Yang Miang Goh et al.

There is growing interest in using safety analytics and machine learning to support the prevention of workplace incidents, especially in high-risk industries like construction and trucking. Although existing safety analytics studies have made remarkable progress, they suffer from imbalanced datasets, a common problem in safety analytics, resulting in prediction inaccuracies. This can lead to management problems, e.g., incorrect resource allocation and improper interventions. To overcome the imbalanced data problem, we extend the theory of accident triangle to claim that the importance of data samples should be based on characteristics such as injury severity, accident frequency, and accident type. Thus, three oversampling methods are proposed based on assigning different weights to samples in the minority class. We find robust improvements among different machine learning algorithms. For the lack of open-source safety datasets, we are sharing three imbalanced datasets, e.g., a 9-year nationwide construction accident record dataset, and their corresponding codes.