Journal
SAFETY SCIENCE
Volume 133, Issue -, Pages -Publisher
ELSEVIER
DOI: 10.1016/j.ssci.2020.104981
Keywords
Artificial intelligence; Bayesian network; Machine learning; Keyword analysis; Incident data; Process safety management; Latent causes
Funding
- NSERC [RGPIN-2018-05170, DGECR-2018-00219]
- SSHRC Partnership Development [PDG-890-2017-0015]
- University internal matching funds
Ask authors/readers for more resources
Process safety management is a framework that shows a company's commitment to safety and risk management. The research applied machine learning and keyword analysis to classify 8199 incident reports, finding that Asset Integrity & Reliability has the greatest impact on reducing incidents.
Process safety management (PSM) is a framework that demonstrates a company's commitment to process safety, a better understanding of hazards and risks, a comprehensive assessment and management of risks, and enhanced learning from experience to improve overall safety and operational performance. Companies often use an incident data reporting system to execute PSM. While companies keep incident data in thousands of reports, rarely do they glean full value in learning from these to prevent and reduce future incidents. To overcome this challenge, this research applied machine learning and keyword analysis to label and classify 8199 incident reports from an oil and gas company into nine groups identified in the latest version of PSM guidelines published by the Center for Chemical Process Safety (CCPS). To converge on an optimal solution, two different Bayesian network techniques (Tabu and hill climbing) were applied. Both methods resulted in the same map, showing that the Total Number of Incidents has the maximum dependency (50%) on Asset Integrity & Reliability; this means focusing resources on this aspect could reduce the total number of incidents by half. Cross correlation analysis (CCA) was also applied, which validated and confirmed this result. This analysis identifies which measures enhance the company's safety management strategy to reduce these latent causes, but also supports critical thinking, enhanced communication, and learning culture to improve organizational safety.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
Article
Engineering, Industrial
Dynamic risk assessment of chemical process systems using the System-Theoretic accident model and process approach (STAMP) in combination with cascading failure propagation model (CFPM)
Hao Sun, Haiqing Wang, Ming Yang, Genserik Reniers
Summary: To maintain continuous production, chemical plant operators may choose to ignore or handle faults online rather than shutting down process systems. However, the interaction and interdependence between components in a digitalized process system are significant, and faults can propagate to downstream nodes, potentially leading to risk accumulation and major accidents. This study proposes a dynamic risk assessment method that integrates the system-theoretic accident model and process approach (STAMP) with the cascading failure propagation model (CFPM) to model the risk accumulation process. The proposed method is applied to a Chevron refinery crude unit and demonstrates its effectiveness in quantifying the process of risk accumulation and providing real-time dynamic risk profiles for decision-making.
SAFETY SCIENCE (2024)
Article
Engineering, Industrial
Modelling a major maritime disaster scenario using the universal modelling framework for sequential decisions
M. Rempel
Summary: This article examines a major maritime disaster scenario and explores the evacuation process in such a situation. The study finds that there are various factors that affect the number of lives saved, including the uncertainty of individuals' medical condition, the arrival time of maritime and air assets, and the decision policies used. The authors formulate the multi-domain operation as a sequential decision problem using a modeling framework and provide decision support through a hypothetical case study.
SAFETY SCIENCE (2024)
Review
Engineering, Industrial
Virtual reality for safety training: A systematic literature review and meta-analysis
D. Scorgie, Z. Feng, D. Paes, F. Parisi, T. W. Yiu, R. Lovreglio
Summary: This study investigates the application and effectiveness of VR safety training solutions in various industries such as construction, fire, aviation, and mining. The findings suggest a need for more studies that adopt theories and measure long-term retention. Two meta-analyses demonstrate that VR safety training outperforms traditional training in terms of knowledge acquisition and retention.
SAFETY SCIENCE (2024)
Article
Engineering, Industrial
Characterization of biases and their impact on the integrity of a risk study
Shital Thekdi, Terje Aven
Summary: This paper examines biases in risk studies and investigates how to identify and address them to ensure high-quality risk analysis. By considering biases related to systematic error, event inclusion, models, and cognitive factors, the paper explores their influence on risk characterization. The insights gained from this exploration can be valuable to risk analysts, policymakers, and other stakeholders involved in risk study applications.
SAFETY SCIENCE (2024)
Article
Engineering, Industrial
A longitudinal study on the impact of occupational health and safety practices on employee productivity
Maryam Lari
Summary: Occupational health and safety (OHS) are crucial for employee well-being and productivity. This study examines the impact of OHS practices on employee productivity in a UAE Fire and Security company, finding that OHS interventions can enhance workplace ambiance and significantly boost employee productivity.
SAFETY SCIENCE (2024)
Review
Engineering, Industrial
Resilience training for critical situation management. An umbrella and a systematic literature review
Elleke Ketelaars, Cyrille Gaudin, Simon Flandin, Germain Poizat
Summary: This systematic literature review examines the literature on resilience training (RT), specifically focusing on the effectiveness of RT interventions in preparing professionals to effectively respond to critical situations. The review identifies five types of RT and suggests the need for conceptual advancements, vocational education and training perspectives, and a cross-disciplinary approach in future research to enhance resilience in safety-related domains.
SAFETY SCIENCE (2024)
Article
Engineering, Industrial
The study of self-organised behaviours and movement pattern of pedestrians during fire evacuations: virtual experiments and survey
Meng Shi, Zhichao Zhang, Wenke Zhang, Yi Ma, Hanbo Li, Eric Wai Ming Lee
Summary: This study investigates pedestrian behaviours and evacuation processes in both fire and non-fire conditions using Minecraft. The results demonstrate the potential of Minecraft for realistically simulating evacuation processes, as the behaviours and flow patterns of pedestrians in virtual experiments fit well with real-life experiments. The study also shows that pedestrians exhibit fire avoidance behaviours and orderly queuing during a fire emergency, resulting in faster evacuation.
SAFETY SCIENCE (2024)
Article
Engineering, Industrial
Riding the wave of sustainability: Integrating OSH into education
Andrea Bikfalvi, Esperanza Villar Hoz, Gerusa Gimenez Leal, Monica Gonzalez-Carrasco, Nuria Mancebo
Summary: This paper proposes a solution for integrating occupational safety and health (OSH) into education, combining theoretical foundations and empirical evidence. The findings include analysis of teachers as stakeholders, barriers and facilitators of OSH integration, and the development of an ICT tool for interaction and sharing in this field. The main contribution lies in envisioning, orchestrating, and validating a solution to integrate OSH into schools and ultimately contribute to sustainable development goals.
SAFETY SCIENCE (2024)
Article
Engineering, Industrial
Critical Entities Resilience Failure Indication
David Rehak, Alena Splichalova, Martin Hromada, Neil Walker, Heidi Janeckova, Josef Ristvej
Summary: This article discusses the adoption of a new directive on the resilience of critical entities and emphasizes the importance of assessing their level of resilience in relation to current security threats. The authors have developed a tool, known as the CERFI Tool, which uses a probabilistic algorithm to predict the failure point of critical entity resilience based on the relationship between threat intensity and protection. The tool is important for increasing the safety of technically oriented infrastructures, particularly in the energy and transport sectors.
SAFETY SCIENCE (2024)
Review
Engineering, Industrial
Machine learning and deep learning for safety applications: Investigating the intellectual structure and the temporal evolution
Leonardo Leoni, Ahmad Bahootoroody, Mohammad Mahdi Abaei, Alessandra Cantini, Farshad Bahootoroody, Filippo De Carlo
Summary: This paper presents a systematic bibliometric analysis (SBA) on the research of machine learning and deep learning in the field of safety. The main research areas, application fields, relevant authors and studies, and temporal evolution are investigated. It is found that rotating equipment, structural health monitoring, batteries, aeroengines, and turbines are popular fields, and there is an increase in popularity of deep learning and new approaches such as deep reinforcement learning.
SAFETY SCIENCE (2024)
Article
Engineering, Industrial
Effects of flight crew role assignment on aviation accidents and incidents: Evidence of a systemic safety issue
Tom Becker, Peter Ayton
Summary: By analyzing global civil aviation data, we found that there is a significant increase in the number of accidents and safety critical incidents, as well as the fatalities, when the Pilot-in-Command acts as the Pilot Flying instead of the Pilot Monitoring. Most of these events occurred in technically airworthy aircraft without any emergencies, and the flight crew assessed them as preventable. These findings align with the crew assignment effect, suggesting that role-dependent status hierarchy and cognitive overload contribute to ineffective flight crew teamwork. The measures implemented to enhance flight crew teamwork, such as Crew Resource Management training, have not been successful in preventing these issues.
SAFETY SCIENCE (2024)
Article
Engineering, Industrial
Audit masquerade: How audits provide comfort rather than treatment for serious safety problems
Ben Hutchinson, Sidney Dekker, Andrew Rae
Summary: This study found that health and safety audits often fail to identify critical deficiencies, with corrective actions mainly focusing on superficial fixes rather than addressing significant operational risks.
SAFETY SCIENCE (2024)
Review
Engineering, Industrial
Smart Personal Protective Equipment (PPE) for construction safety: A literature review
Sina Rasouli, Yaghoub Alipouri, Shahin Chamanzad
Summary: Construction projects are risky environments, but the development of Personal Protective Equipment (PPE) and comprehensive safety management can effectively control the number of accidents.
SAFETY SCIENCE (2024)
Article
Engineering, Industrial
Is drug driving more common among those who know where the police are? An investigation into the use and non-use of Facebook police location communities
Laura Mills, Verity Truelove
Summary: This study investigated the use of police location communities (PLCs) for obtaining information about roadside drug testing (RDT) among drivers, and found that drivers who used PLCs were more concerned about being caught for drug driving, had a better understanding of the related penalties and procedures, and knew others who also used PLCs to avoid detection for drug driving. Furthermore, the study found that the use of PLCs was associated with choosing back roads for driving, which may reduce the risk of detection for drug driving.
SAFETY SCIENCE (2024)
Review
Engineering, Industrial
Glimpse of safety science development in China: A review of safety fundamental research and construction of six new postgraduate courses for safety majors by safety & security theory innovation and promotion Center of Central South University
Chao Wu, Xi Huang, Bing Wang
Summary: After decades of development, the safety discipline in China has made remarkable progress. The project led by the Safety & Security Theory Innovation and Promotion Center of Central South University has filled the gaps in safety science education by creating textbooks and courses for postgraduate students. These achievements have played an important role in the development of safety science and can serve as a reference for basic research and talent training in safety science globally.
SAFETY SCIENCE (2024)
Share Paper
