A novel job rotation schedule model for repetitive jobs regarding posture variety during work hours and solving with a genetic algorithm

Abstract

Job rotation is a known method that is often used to reduce monotonous workloads on workers with repetitive workstation-based jobs. Changes in a worker’s body posture can contribute to reduce the monotony; particularly, while there exists none or only minimal external force exertion. The purpose of this research is to develop a method to incorporate posture variety, individually, for each particular body area, into the rotation. This method can increase the possibility of having overall posture variety during work-hours or shift-by-shift for workers. To this end, fuzzy dissimilarity magnitudes between two jobs based on linguistic variables are defined and then used to propose new criteria. According to the criteria, an integerprogramming model for the rotation is developed. Owing to the large search space in which to find a very good solution (approximated optimum solution), a conventional genetic algorithm and a customized cellular genetic algorithm are employed and compared. In addition to being intuitively logical, the algorithms are examined in a simplified test case with six different assembly jobs (performing assigned tasks repetitively), and the results indicate that the cellular genetic algorithm can efficiently find better job rotation schedules to satisfy the criteria.