An Agent-Based Modeling Approach for Understanding the Sociocognitive Process of Construction Workers’ Unsafe Behavior

Abstract

Many accident investigations reveal workers’ unsafe behavior are the major and frequent causes of on-site accidents, and thus need to be addressed urgently. Behavior is the result of cognition. Using a cognitive perspective can reveal the essential reasons for the choice of construction workers’ unsafe behavior. Workers’ unsafe behavior is not only the result of isolated individual cognitive processes, but also the result of their interaction with others. Therefore, taking both individual cognitive psychology and social factors into consideration, this paper aims to develop an Agent-Based Model of construction workers’ sociocognitive processes simulated with MATLAB software. Taking a certain engineering project as the research object, site questionnaires and surveys were implemented for the design of simulated working environment and individual attributes. The developed model is applied to analyze how the workers’ sociocognitive process interact with coworkers, foreman and managers. The results indicated that (1) under sociocognitive level, the degree of workers’ risk acceptance is lower than that of individual level, and the number of unsafe behaviors is less; (2) Workers under the influence of the foreman have the lowest risk acceptance; (3) proper management measures and high social identity can reduce the proportion of workers’ unsafe behavior. Based on the above results, managers should proceed from the workers’ sociocognitive process and adopt appropriate management methods to reduce unsafe behavior of workers.

Keywords: construction worker; unsafe behavior; sociocognitive process; agent-based modeling