Construction of a Bayesian network model for improving the safety performance of electrical and mechanical (E&M) works in repair, maintenance, alteration and addition (RMAA) projects

• Published in: Safety science 2020-11, Vol.131, p.104893, Article 104893
• Main Authors: Chan, Albert P.C., Wong, Francis K.W., Hon, Carol K.H., Choi, Tracy N.Y.
• Format: Article
• Language: English
• Subjects: Accident analysis; Bayesian analysis; Bayesian networks approach; Construction; Electrical and mechanical (E&M) Works; Inspection; Maintenance; Mechanical engineering; Occupational safety; Repair; Safety; Safety factors; Safety management; Safety measures; Smoking
• ISSN: 0925-7535; 1879-1042
• DOI: 10.1016/j.ssci.2020.104893

•Alcohol and smoking habits of workers exert a considerable influence on the safety performance.•Working experience of workers had the least impact on improving safety performance.•Joint strategies with controlling more than one factor further improve safety performance.•The improvement percentage of safety performance follows a decreasing trend. The safety concern and volume of repair, maintenance, alteration and addition (RMAA) works have significantly increased in recent years. RMAA works include a variety of work trades. Electrical and mechanical (E&M) works are regarded as one of the most hazardous trades with numerous complex activities. However, the research on the safety of E&M works in RMAA projects is limited. This study aims to develop a Bayesian network (BN) model that encapsulates the interrelationships between safety factors and safety performance. Survey data are analysed with factor and BN analyses to construct a BN model. Findings show that alcohol consumption and smoking habits of workers exert a considerable influence on the safety performance of workers. A strategy via controlling multiple factors (joint strategies) may even improve safety performance. Analytical results indicate the effectiveness of a joint control of alcohol and smoking habit, safety inspection and procedures factors would be the most effective strategy to improve safety performance. The significance of this study lies in the proffering of a BN model that reveals the interrelationships of the safety factors and safety performance of E&M works in RMAA projects. The findings will help in formulating effective safety management strategies to improve the safety of RMAA works. The BN model can be a practical technique to diagnose effective safety measures for improving safety performance. The research outcomes would be valuable to key project stakeholders of E&M works to achieve better safety performance and bring tremendous value in better safeguarding E&M workers’ health and safety.