Assessing the Off-Site Manufacturing Workers’ Influence on Safety Performance: A Bayesian Network Approach
Sadith Chinthaka Vithanage, Michael C.P. Sing, Peter Davis, Mohammad Tanvi Newaz
Abstract
Off-site manufacturing (OSM) offers a wide range of benefits to the construction industry by saving time, reducing waste, being an environmentally friendly solution, and providing a much safer onsite environment. However, safety management of OSM has become a concern due to the worker-related safety issues, which increase the safety incidents in off-site factories. However, research on the safety of OSM activities within a factory environment are still very limited. Therefore, this study aims to examine the interrelationships among worker-related safety factors and their influence on OSM safety performance. A probabilistic model based on Bayesian networks (BNs) to assess the influence of worker-related safety climate factors on OSM safety performance is utilized in this research study. A comprehensive review and evaluation of causal safety factors with the support of a questionnaire survey with OSM industry practitioners are the basis for this BN model. The proposed BN model is then verified by conducting a sensitivity analysis such as tornado diagrams and derivatives of sensitivity. The established model presents probabilities associated with different states of safety factors and identifies the interrelationships among worker-related safety climate factors and OSM safety performance. The research findings show that improvements in management safety response, coworkers’ safety values and practices, and quality of training exert a greatest effect on the safety performance. Analysis further indicates that a balance between safety and production significantly affects workers’ safety knowledge. This study contributes to the OSM safety domain by offering an effective tool to predict safety performance.