Understanding the Effect of Management Factors on Construction Workers' Unsafe Behaviors Through Agent-Based Modeling

As the information technology and its application level improve, the construction industry is transforming and upgrading from traditional construction to smart construction in recent years. However, safety issues still exist and construction accidents occur frequently. Construction workers' unsafe behaviors are the primary reason for accidents; how to reduce workers' unsafe behaviors by conducting safety management is an urgent and crucial subject. The study considered the intelligent video surveillance and five traditional management modes into the management system and proposed an advanced cognitive model of construction workers' unsafe behaviors. The agent-based modeling method was adopted to explore the action mechanisms of typical management factors on the cognitive attributes of workers' unsafe behaviors. A case study was undertaken based on a prefabricated residential building construction project, and the theoretical model was verified by comparison with empirical data. Sensitivity analysis was conducted to determine the relative importance of different management factors. The results indicate that managers' positive demonstration roles and behavior feedback have the best performance on preventing the spread of unsafe behaviors; daily meeting, appropriate frequency of communication, and all-directional intelligent video surveillance can effectively reduce unsafe behaviors, while frequent safety inspection has little effect on improving the safety performance. Connecting with the present development situation of construction industry, these findings are beneficial as an academic reference for transforming construction enterprises to formulate effective safety management strategies and prevent accidents.