An Agent-Based Framework for Investigating Safety-Productivity Tradeoff of Construction Laborers Considering Risk-taking Behavioral Heterogeneity

Construction laborers and crews play a critical role in achieving a safe and productive construction site. Many past research studies used top-down approaches/perspectives for studying the impact of laborers' performance on overall construction site outputs with limited flexibility in accounting for laborers' various characteristics. However, the recent reap in computational advances allowed applications of bottom-up architectures, which can potentially incorporate heterogeneous characteristics of laborers' individual behavioral and decision-making features effectively. Accordingly, agent-based modeling (ABM), as a tool to leverage a bottom-up methodological approach, has been widely adopted by recent research. Existing literature investigated the influence of changes in laborers' behaviors and interactions on either construction sites' safety performance or productivity performance individually, leaving the tradeoff between safety and productivity in this context relatively unexplored. Accordingly, this study aims to develop an agent-based framework to study the tradeoff between project safety and productivity performances resulting from changes in laborers' behaviors after attending safety trainings. Our findings via simulations indicate that proper safety trainings can improve safety performance without negatively impacting productivity performance.