A dispatching algorithm for flexible job-shop scheduling with transfer batches: an industrial application

In this study, we present a dispatching algorithm to solve a real-world case of the flexible job-shop scheduling problem with transfer batches and the objective of minimising the average tardiness of production orders. The proposed algorithm considers two variants: (i) an ordered variant, where the priority dispatching rules are applied in a predefined order, and (ii) a randomised variant, where the user can assign probabilities (weights) to the priority rules. Using the information of the number of units and due dates requested by the only customer, the algorithm provides the sequence of operations that must be performed on each machine, as well as the start and completion times of operations. In order to reduce the impact of unexpected events on a generated schedule, several robustness rules are considered. The obtained results substantially improve the former method used in the company in terms of minimising average tardiness. Additionally, other important benefits are obtained, including significant saving in the time spent on scheduling, simplicity of use of the proposed procedure, robustness against unexpected events, reduction of idle times, improvement of decision-making information, and improvement of on-time delivery performance.