Modeling and analyzing workers' hazard perception in construction equipment operations: A hierarchical Bayesian cognitive modeling approach integrating multisource data

Hazard perception is essential for workers engaged in continuous construction equipment operations within dynamic, complex environments. Existing research lacks a comprehensive understanding of how internal and external factors, along with a series of cognitive processes, shape the uncertainty of hazard perception. Consequently, there is limited insight into hazard perception performance and associated error risks among different workers in various task situations. This study proposes a hierarchical Bayesian cognitive modeling approach integrating multisource data to explore the cognitive mechanisms underlying workers' hazard perception and quantify corresponding errors. A cognitive model represents how workers perceive and respond to hazards, whereas a hierarchical Bayesian model quantifies multilevel dependencies between cognitive characteristics and hazard perception outcomes. The proposed modeling approach incorporates mental fatigue and individual differences through hierarchical priors that distinguish task temporal variations and worker-specific traits. The Markov chain Monte Carlo method estimates posterior distributions of cognitive parameters, capturing the uncertainty in hazard perception. The data from a behavioral experiment demonstrated the feasibility of the approach. The results confirmed that hazard perception is a continuous, intricate cognitive process and revealed that relevant cognitive parameters are significantly influenced by mental fatigue and individual differences. The approach integrates multisource data, captures time-varying cognitive uncertainties, and examines key factors affecting hazard perception. This method exhibits strong potential for real-time identification and prediction of hazard perception errors across different workers, ultimately enhancing targeted management of workers' hazard perception performance in intricate construction equipment operations, improving safety management judgments, and mitigating risks.