Reactive Flexible Job Shop Problem with Stress Level Consideration
Abstract
Taking into account a real-world issue, the present study focuses on a flexible job shop scheduling problem (FJSSP) that deals with new job arrivals. This problem is very common in real-world manufacturing operations. On the other hand, Industry 5.0 environment pays more attention to human resources, and it is shown that well-being of workers including the stress level of them has a great impact on shop scheduling performance performance. However, with the arrival of a new job, the initial planning needs a rescheduling and these changes on the initial schedule may increase the stress level of workers. Still, given the real-world problem, we want to minimize the stress level of different rescheduling strategies. Three types of changes will be imposed to the shop floor schedule, which could lead to an increased stress level on human resources. These changes are as follows: 1 - Shifting an old operation on the same machine when the new job arrives; 2 - changing the machine assigned to an operation; 3 - altering the sequence of the operations.
The procedure for calculating the different kinds of changes affected by the new job arrivals is illustrated in order to find the level of imposed stress. To solve such an NP-Hard problem, a Genetic Algorithm (GA) is investigated to solve it. At first an initial schedule is built based on benchmark instances from FJSSP literature. Then, at different times, new jobs will arrive, with routes taken from the same instance. The instances are used to validate the proposed algorithm. Behavior of the algorithm on very large problems indicates that obtained schedules remain as compact as expected while considering the stress criterion.