Agent based simulation of workers' behaviours around hazard areas in manufacturing sites

El-Raoui, Hanane and Quigley, John and Sherlock, Andrew and Aslan, Ayse and Corney, Jonathan and Hanson, Jack and Vasantha, Gokula; (2023) Agent based simulation of workers' behaviours around hazard areas in manufacturing sites. In: Proceedings of the Simulation Workshop 2023. OR Society.

[thumbnail of El-Raoui-etal-ORSW-2023-Agent-based-simulation-of-workers-behaviours-around-hazard]
Preview
 Text. Filename: El_Raoui_etal_ORSW_2023_Agent_based_simulation_of_workers_behaviours_around_hazard.pdf
Accepted Author Manuscript
License: Strathprints license 1.0
Download (578kB)| Preview

Abstract

Rewards for risk taking behaviour by workers (if accidents do not occur) can be realised in the form of increased productivity or worker idle time. However, frequent unsafe behaviours of workers inevitably results in accidents and an associated loss in productivity. Workers' behaviour towards safety is influenced by management, who can encourage or discourage risk taking behaviour. In this paper, we explore the relationship between the perceived reward by individual workers who expose themselves to hazards and a management response in the form of inspections to monitor and address inappropriate behaviours. We conduct this study by developing an Agent Based Simulation Model, where workers are required to learn paths within a factory exposed to hazardous areas, with inspectors randomly moving around the factory to correct inappropriate behaviour if noticed. We assume workers are maximising their anticipated reward as they learn routes through the factory. This agent based model is used to explore the impact of inspection frequency and reward perception (i.e. parameters which can be influenced by management) on the number of workplace accident. The results demonstrated that the proposed model is a valuable tool to assist the management in predicting the potential safety improvement from safety management practices focusing on safety inspections, and changing workers perceptions.

CORE (COnnecting REpositories)