A Genetic Algorithm to Design Job Rotation Schedules with Low Risk Exposure

Abstract

In automotive industries, the manufacturing processes are characterized by repetitive tasks and physically demanding work, with possible long-term implications on the musculoskeletal health of the workers. One key organizational strategy that provides an improvement in the prevention of musculoskeletal disorders are job rotation schedules. These are usually designed manually, being (1) time demanding and (2) a subjective evaluation of the schedule’s risk. In this work, a genetic algorithm is presented, to generate automatically a daily job rotation schedule. The quality of the schedule is based on objective scores as- signed to workstations by the European Assembly Worksheet (EAWS) risk screening tool, guiding the algorithm in reaching a final solution that promotes schedules with lower exposure to the sequence of workstations assigned to each worker of the team. The schedules generated by the algorithm were compared to schedules designed by the team leaders and presented a better overall result.