Industrial accidents in chemical process engineering continue to pose a significant issue despite the widespread use of Industry 4.0 technology and data-driven monitoring systems. Traditional safety frameworks often depend on either purely empirical machine learning models or deterministic first-principles simulations, creating a methodological split that constrains prediction reliability in uncommon, high-impact situations. This work bridges the structural gap by incorporating physics-informed artificial intelligence into a digital twin architecture for the avoidance of industrial accidents. A methodological framework driven by simulation was established, integrating first-principles process modeling, synthetic data generation with controlled fault injection, supervised and unsupervised learning, and reinforcement learning for safety-constrained optimization. Physics-based limitations were included into the learning aim to maintain thermodynamic and transport consistency. The model's performance was assessed using safety-oriented criteria, such as detection delay, false negative rate, resilience to sensor noise, and stability amid parametric uncertainty. Results demonstrate that physics-informed models significantly reduce detection latency and false negatives in accident precursor regimes compared to purely data-driven baselines. The integration of constraint-aware learning improves extrapolation stability under class imbalance conditions typical of industrial safety datasets. Furthermore, explainable AI mechanisms enhance interpretability and regulatory transparency. These findings indicate that AI-enhanced simulation models reconfigure accident prevention strategies by shifting from reactive threshold systems to proactive, mechanism-consistent risk anticipation frameworks applicable to safety-critical chemical processes.
Wassim Dbouk, D. Teagle, Lindsay-Marie Armstrong
et al.
Carbon Capture and Storage (CCS) is an essential component of the UK Government ’ s net-zero strategy. Policies emphasize the need for flexible and accessible CO ₂ transport and storage networks, with shipping emerging as a key non-pipeline transport modality to connect industrial clusters to offshore storage. In this article, we assess whether current health and safety and major accident prevention regulations adequately govern the risks posed by expanding CO ₂ handling and storage in UK ports to support CCS deployment. Our analysis identifies three regulatory gaps. First, while the Port Marine Safety Code addresses regulatory complexity in UK ports through establishing uniform national standards for marine safety, it cannot regulate the emerging risks of anticipated large-scale CO ₂ shipping activities without clear performance standards in specific legislation. Second, duly appointed harbor masters must be well-informed to effectively exercise the powers granted under the Dangerous Goods in Harbour Areas Regulations (DGHAR) to reduce serious accident risks associated with increased CO ₂ shipping. Third, the Control of Major Accident Hazards Regulations (COMAH) currently exclude temporary CO ₂ storage and do not include CO ₂ within their scope, limiting their effectiveness for major accident prevention in port storage scenarios. To address these gaps, we recommend issuing tailored guidance under DGHAR to clarify risk management responsibilities for CO ₂ shipping and amending COMAH to include CO ₂ storage and recognize CO ₂ as a dangerous substance. These reforms are essential to protect port communities, ensure robust risk management, and support the safe, sustainable expansion of CO ₂ shipping as a critical enabler of CCS.
Ali Nasiri, Esmaeil Salimi, Morteza Delfan Azari
et al.
Flood zoning has extensive applications in flood management and is considered one of the fundamental and critical pieces of information in flood risk management. Flood zoning in urban areas is much more challenging than modeling in floodplain and river areas due to the two-dimensional nature of the flow and, on the other hand, the density of urban features such as buildings, streets, boulevards, and public pathways. In this study, flood zoning for districts 21 and 22 of Tehran was conducted under the current conditions, where the area is almost devoid of surface water collection channels, using a physically-based rainfall-runoff model and two-dimensional hydraulic routing which is the novelty aspect of the article. For this purpose, the HEC-HMS model was used to estimate the runoff from the mountains, and the MIKE model was used to simulate urban rainfall-runoff. According to the modeling results, the areas affected by a 50-year flood event were identified using an integrated modeling approach in districts 21 and 22, covering 8% of these areas. In these areas, the maximum flood depth is 11.8 meters in Vardavard river and the highest speed is 4.5 meters per second at the beginning of Hashemzadeh street (south of Kharrazi highway). The results indicate that in the event of extreme events such as a 50-year rainfall, a significant portion of the highways and main communication arteries of Tehran leading westward would be disrupted, and traffic would be impossible. Moreover, various land uses would fall within the flood zone, and due to the absence of a surface water network, waterlogging conditions throughout districts 21 and 22 of Tehran are predictable. Therefore, the development of a surface water collection network is one of the main priorities for reducing flood risk in these areas.
Risk in industry. Risk management, Industrial safety. Industrial accident prevention
Kateryna Hutchenko , Vyacheslav Kozachuk, Oleh Hutchenko
et al.
Одним із проблемних питань, що набуло актуальності під час російсько-української війни перед Збройними Силами України, є підвищення ефективності функціонування системи реабілітації військовослужбовців Збройних Сил України. Така система має відповідати науково обґрунтованим вимогам. Тому виникла необхідність вирішити завдання, що пов’язане з побудовою відповідної математичної моделі.
Мета статті. Розроблення методичного підходу до побудови моделі системи реабілітації військовослужбовців Збройних Сил України.
Методи дослідження. Для безпосереднього моделювання системи реабілітації військовослужбовців застосовано математичні апарати теорії масового обслуговування, зокрема, мова йде про апарати систем масового обслуговування і мереж масового обслуговування.
Отримані результати дослідження. Результати аналізу досліджень і публікацій дають змогу стверджувати, що моделям, які стосуються реабілітації військовослужбовців Збройних Сил України надається достатньо уваги. Загалом ці моделі містять показники виконання одного із завдань, а безпосередньо призначених для характеристики параметрів системи реабілітації особового складу та її продуктивності, немає. Зважаючи на викладене, постає нагальне проблемне питання, в побудові моделі системи реабілітації військовослужбовців, удосконаленні її показників, які б враховували продуктивність надання медичної допомоги і можливості підрозділів із реабілітації військовослужбовців Збройних Сил України. Для безпосереднього моделювання системи реабілітації військовослужбовців застосовано математичні апарати теорії масового обслуговування. Зокрема, мова йде про апарати систем масового обслуговування і мереж масового обслуговування. Ґрунтом для моделювання слугували прийняті зараз у Збройних Силах України порядок, способи та підходи до побудови системи реабілітації військовослужбовців, а також алгоритми її дії. Сучасну систему реабілітації (медичну, фізичну та психологічну реабілітації) військовослужбовців можна уявити як мережу масового обслуговування, яка має низку особливостей і для розгляду якої зроблені деякі припущення. У цій науковій роботі, використовуючи вищенаведені методи дослідження, були розраховані параметри системи реабілітації військовослужбовців, яка діє в Збройних Силах України сьогодні.
Елементами наукової новизини є комплексне рішення завдання формування системи реабілітації, застосовуючи модель системи реабілітації, створеної на основі підходів теорії мереж масового обслуговування.
Теоретична й практична значущість викладеного зводиться до створення моделі багаторівневої системи, яка дає можливість розраховувати систему в цілому та її фрагменти. Перспектива подальших досліджень: застосування запропонованого підходу для визначення можливості використання розімкнутої мережі масового обслуговування у випадках, коли тривалість обслуговування заявок у всіх вузлах являє собою випадкові величини, які розподілені не за експоненціальним, а за іншими законами.
This paper presents the design ultrasonic sensor-based work holding system aimed at significantly enhancing industrial safety and accident prevention in machining and fabrication environments. Traditional work holding devices, such as vices and clamps, rely on manual inspection and mechanical force, which can be prone to human error, leading to improper workpiece seating, slippage, and catastrophic tool or part failure. The proposed system integrates an array of ultrasonic sensors to non-invasively and continuously monitor the proximity, secure seating, and potential movement of the workpiece relative to the holding jaws. A microcontroller processes the sensor data, providing real-time feedback to the machine operator and, crucially, generating a mandatory interlock signal to prevent machine operation unless the workpiece is confirmed to be held securely within pre-defined safety tolerances. This intelligent system minimizes the risk of accidents caused by improperly secured workpieces, promotes a safer working environment, and contributes to enhanced operational efficiency and compliance with safety regulations.
Accidents are a constant problem in numerous enterprises and they greatly affect workers and project results. This study aims to find out what caused these crashes, focusing on how worker-related, environmental and managerial factors all interact with each other. The goal is to find the main factors affecting Workplace Hazard Prevention (WHP) and make a model that can predict the future to lower risks. This study uses an ensemble machine learning (EML) approach to show Industrial Accident Analysis and Predictive Models for Workplace Hazard Prevention (IAA-PM-WHP). An analysis is conducted on a publicly accessible collection of 65,518 workplace injury reports from the Occupational Safety and Health Administration (OSHA), using four distinct ML models. This study suggested a way to build a model that takes into account three important factors: "type of damage," "kind of event," and "harmed organ." The EML model integrates predictions from four fundamental ML methodologies via soft voting. Among classic ML models, the RF method had the greatest accuracy (0.89), indicating robust overall prediction power. The EML method outperformed all models, attaining the greatest accuracy (0.92), precision (0.99), recall (0.899), F1-score (0.94) and AUC (0.92).
One of the major challenges faced by the manufacturing industry is the prevention of workplace accidents. In this context, ensuring compliance with safety regulations, such as the use of personal protective equipment (PPE) and the proper execution of specific tasks according to safety protocols, is essential, especially when supervisors or safety personnel are not present on site. To address this issue, we propose a multimodal method for real-time human action prediction in industrial environments, aimed at supporting accident prevention systems. Our approach integrates two parallel Graph Attention Networks (GATs): one based on human skeleton pose estimation, and another built from scene object detection graphs. By combining these two complementary modalities, the model captures both human motion dynamics and contextual environmental information. To the best of our knowledge, this approach has not yet been explored in the literature. The proposed method will be evaluated on two benchmark datasets: Kinetics-400 (a large-scale video dataset with diverse real-world actions), and UnsafeNet (a dataset featuring factory-recorded videos annotated with safe and unsafe behaviors). The expected results aim to demonstrate the feasibility of applying multimodal GAT-based architectures to enhance occupational safety through intelligent action recognition systems.
Due to advancements in the fields of machine learning and computer vision, compliance monitoring systems for industrial safety by enterprise workers have become widely adopted in automating the solution of various classes of tasks. The focus of this work is on automating the monitoring of safety violations to prevent accidents. The specific task of this work is to monitor compliance with safety regulations when moving on stairs. An important direction in solving such tasks is the development of computer vision systems aimed at detecting human actions. This work proposes using a neural network for human pose estimation as a feature extraction network. A distinctive feature of the proposed solution is the implementation of a decision-making algorithm based on an LSTM network with a small number of parameters, and the use of optimized decisionmaking methods for operation on devices with limited computational power. All utilized neural networks were exported to TensorFlow Lite format to enhance performance when running on a central processing unit. The highest achieved accuracy was 98.2%, with a precision of 97.8% and a recall of 98.5% at a confidence threshold of 0.5.
The scenario of industrial safety has been shifting significantly within the last 50 years, with the meaning of which is the transition to the design that is inherently safe (ISD) as the most effective tool for preventing hazards. It has been stated statistically that industrial accidents can be reduced by as much as 85 percent through the systematic delivery of the principles of ISD, which are referred to as minimization, substitution, moderation, and simplification. In this paper, we are taking a historical look at the development of ISD since the post-Flixborough period, its integration into the contemporary regulation, and the interaction of technology, human factors, and safety performance. There is a focus on measurable performance indicators, financial value, and life-cycle principles of implementation, through the facility life stages. High technologies, including digital twins, real-time monitoring systems, and predictive analytics, are mentioned as the factors contributing to a more rapid adoption of ISD, whereas human factors engineering is mentioned as one of the key factors ensuring sustained operations safety. In the findings, it was confirmed that the ISD actually enhances the implementation of process safety and regulatory compliance, besides bringing huge savings, lower insurance premiums, and enhancing the emergency process both in terms of cost and time. This study highlights the timelessness of the mission to eradicate hazards at the source as the only foolproof method of ensuring industrial processes are not jeopardised in the age of Industry 4.0 and beyond.
The Seveso directives on the control and prevention of major accidents prescribe the obligations of industries that contain hazardous substances in specified quantities. To meet the requirements for safe technology, it is necessary to conduct complex field research, identify potential accidents, perform simulations of accident scenarios, and model their effects, consequences, vulnerable zones, and risk assessment. This paper presents the results of research at the location of a Seveso operator (Contract: JN122OU/15) and simulates a complex case of the worst possible accident, specifically a terrorist attack on fuel oil tanks, resulting in an explosion that would spread to the zinc-oxide warehouse through a domino effect. The computer program ALOHA was used to model the dispersion of toxic gases, determine the consequences of explosions, analyze vulnerability, and identify danger zones to obtain relevant data for the development of protection measures and accident response
PurposeDespite efforts to improve safety management practices in industrial companies, major accidents seem to be inevitable. Many accidents still occur because companies are unable to learn from past occurrences due to ineffective incident and accident learning processes. This study proposes a learning-based framework for industrial accidents investigation and contributes to accident prevention research.Design/methodology/approachThe proposed learning process includes the analysis of the industrial accident using the Event Tree Analysis (ETA) method, capitalisation of causative factors using the Swiss Cheese Model (SCM), and finally modelling the relationships among the accident causative factors and analysing their causality using the Fuzzy Cognitive Mapping (FCM) technique and running learning scenarios.FindingsThe proposed learning process was applied to an industrial accident, and the results showed that human unsafe behaviours and unsafe supervision were the principal causative factors of the blowout accident.Practical implicationsThe proposed learning-based framework provides a structured approach for oil and gas companies to systematically analyse and learn from past accidents, enhancing their prevention strategies. Theoretically, the framework bridges the gap between theory and practice by demonstrating how established accident analysis methods can be combined and applied in a real-world industrial context.Originality/valueThe proposed learning process combines accident analysis and investigation techniques with simulations for an in-depth and robust learning-based framework for accident prevention.
Ju-Han Song, Seung-Hyeon Shin, Sungmin Kang
et al.
With increasing industrial sophistication and complexity, workplaces are increasingly prone to occupational accidents, causing negative impacts on workers and employers, including economic losses and decreased productivity. South Korea occupational safety and health has implemented new policies addressing potential risks to overcome stagnation in industrial accident reduction and predict site accidents from past cases. Cases are human-classified according to rules, including occurrence type or original causal materials. However, human errors, subjective judgments, synonyms, and terms incorrectly used by classifiers reduce original data quality and impede developments or applications of policies, technologies, and methods preventing accidents based on past accidents. This study proposes three artificial intelligence models to objectively classify the occurrence type of accident cases. Models are developed based on a natural language processing model (KoBERT), which considers Korean language characteristics. Each model is tested by sequentially performing sentence preprocessing, keyword replacement, and morphological analysis. The proposed Model 3 exhibits 93.1% accuracy, which was the highest among tested models. Up to three classification categories for occurrence type are allowed to assist objective classification. The accident case-based occurrence type classification model is effective for industrial accident prevention, aiding in strategy development and reducing social costs.
With the rapid progress of industrialization, serious accidents have frequently occurred at industrial sites. Despite the existence of several laws and regulations for preventing industrial accidents, including the Occupational Safety and Health Act, social criticism has been raised regarding their inefficiency. To address this issue, the Serious Accident Punishment Act was implemented. The Ministry of Employment and Labor and the Korea Safety and Health Agency have prepared and implemented various industrial accident prevention measures and programs; however, the number of industrial accidents has not decreased. Therefore, new evaluation criteria must be introduced when establishing policies or systems to prevent industrial accidents. This study introduces evaluation items used in Europe and the United States as industrial accident prevention items for assessing industrial accident prevention projects and proposes a plan to select priorities for these items by applying the fuzzy TOPSIS technique. Prioritization was determined by calculating four evaluation criteria and ten industrial accident prevention items using a 7-point linguistic scale. Our results allow for the simultaneous consideration of several industrial accident prevention items. Thus, various aspects of industrial accident prevention activities can be comprehensively evaluated, enhancing the effectiveness and execution of safety and health management, thereby helping to reduce industrial accidents.
Abstract Injury prevention during physical education lessons is essential to ensure the safety of students and to enable them to actively participate in sports activities without risk. The term prevention is commonly used to stop or prevent from happening, but it also means to slow, hinder, or prevent an event before it happens. To prevent it is enough to decrease the probability of the occurrence of events from which such accidents result.
This paper outlines the specific provisions of Italian legislation regarding workers’ exposure to electromagnetic fields (EMFs) from 0 Hz to 300 GHz compared to the minimum health and safety requirements set in European Directive 2013/35/EU. In particular, the path to be followed to assess and manage occupational exposure to EMFs is outlined in relation to the distinction between ‘professional’ and ‘non-professional’ exposure of workers, as well as to the precautionary limits regarding exposures from power lines (50 Hz) and broadcast and telecommunication fixed systems (100 kHz–300 GHz) established by Italian regulations. The reasons underlying such an approach—mainly relying on the intent to reconcile scientific evidence with risk perception in public opinion—are analysed and discussed with the aim of increasing the knowledge of national regulatory provisions on occupational risk assessment, which may be more stringent than the requirements envisaged by international guidelines and community regulations.
Industrial safety. Industrial accident prevention, Medicine (General)
Sentagi Sesotya Utami, Winny Setyonugroho, Moch Zihad Islami
et al.
Introduction: Ship-to-shore (STS) crane operators strive for efficiency in their work, but they must take a hard look at their high-risk jobs. It is necessary to learn how to improve occupational safety and health. This study aims to investigate the problems faced by STS crane operators working in container ports and to understand the importance of fit-for-work monitoring procedures, particularly for individuals working in high-risk industries such as STS operators. Methods: This study used a qualitative approach, and data were collected through interviews and observations of STS operators and in-house clinic staff. Nine STS operators, two in-house clinic staff, and two safety, health, and environment (SHE) staff were interviewed. Results: This study found that container terminal companies emphasise two critical aspects for STS operators: productivity and occupational safety and health. STS operators face health problems, including physical and psychological problems, due to the fast-paced work system, sleep patterns, daily activities, and thoughts that are difficult to control. Employees have coping mechanisms to deal with fatigue, and stakeholders have effectively communicated the company's safety and health culture. Most stakeholders in a container terminal company want a fit-for-work monitoring system to make the business efficient and sustainable. Conclusion: The STS industry faces a significant problem with operator fatigue, which can negatively impact safety and productivity. This issue requires a comprehensive strategy, including legislation to regulate working hours and shift patterns, technology to combat fatigue, and operator education and training.
Driver mental fatigue is considered a major factor affecting driver behavior that may result in fatal accidents. Several approaches are addressed in the literature to detect fatigue behavior in a timely manner through either physiological or in-vehicle measurement methods. However, the literature lacks the implementation of hybrid approaches that combine the strength of individual approaches to develop a robust fatigue detection system. In this regard, a hybrid temporal approach is proposed in this paper to detect driver mental fatigue through the combination of driver postural configuration with vehicle longitudinal and lateral behavior on a study sample of 34 diverse participants. A novel fully adaptive symbolic aggregate approximation (<i>faSAX</i>) algorithm is proposed, which adaptively segments and assigns symbols to the segmented time-variant fatigue patterns according to the discrepancy in postural behavior and vehicle parameters. These multivariate symbols are then combined to prepare the bag of words (text format dataset), which is further processed to generate a semantic report of the driver’s status and vehicle situations. The report is then analyzed by a natural language processing scheme working as a sequence-to-label classifier that detects the driver’s mental state and a possible outcome of the vehicle situation. The ground truth of report formation is validated against measurements of mental fatigue through brain signals. The experimental results show that the proposed hybrid system successfully detects time-variant driver mental fatigue and drowsiness states, along with vehicle situations, with an accuracy of 99.6% compared to state-of-the-art systems. The limitations of the current work and directions for future research are also explored.
Industrial safety. Industrial accident prevention, Medicine (General)