Investigation into the determinants of occupational safety for achieving zero accidents in multi-storey building construction projects: A Bayesian belief network approach.

The Indonesian construction sector is vital for economic development but faces high accident rates because of inadequate occupational health and safety (OHS) measures. This study examines causal factors in high-rise construction accidents, focusing on human factors, organization, work environment, and payment systems. A Bayesian Belief Network (BBN) was used to evaluate OHS implementation and identify key elements for achieving zero accidents. Data were collected through a structured questionnaire completed by 28 experts assessing 277 workers across 33 critical safety variable relationships. Analysis using RStudio's bnlearn revealed that regulatory requirements strongly influence safety design, corporate culture, and OHS management, while overly complex safety designs hinder supervision and training. Findings show that workers' physical and mental conditions are vital for zero accidents, whereas safety awareness alone is insufficient. Improved training enhances payment systems, boosting motivation and productivity. The study recommends simplifying safety design and refining OHS procedures to reduce accident risks. The main advantages of the proposed method are:•Using Bayesian Belief Network to analyse work accidents at high-rise projects.•Examined the human, organisational, work environment, and payment system aspects.•Found that simplifying safety design and improving the training are important for achieving zero accidents.