Designing a virtual warehouse operator integrating fatigue, recovery, and learning using agent-based modeling

This study proposes a virtual agent using agent-based modeling to simulate operators’ behavior and sensitivity to their environment. The goal was to develop and calibrate a human-like agent that can be used for future work in distribution center environments using agent-based simulation. By developing a human-like agent, productivity and accuracy were analyzed by considering various environmental control variables. The methodology involved selecting relevant factors, researching mathematical models, collecting data, calibrating and validating the models, and developing a simulation model. The results of the simulation model analyzed the impact of changes in schedules, experience level, task intensity, rest quality, and layout on operator productivity and accuracy. The findings revealed that fatigue and experience significantly influence operators, with factors such as task intensity, rest quality, layout, and schedules affecting their performance. The results used data collected from a real distribution center operation of a Portuguese retailer.