Decision support for the integrated flexible job shop scheduling and operator timetabling problem

This paper addresses the integrated Flexible Job Shop Scheduling and Operator Timetabling (FJS-OT) problem, which incorporates shift-based constraints reflecting operators' timetabling regulations. To support practitioners in making informed decisions, we first enhance the existing mixed-integer programming (MIP) formulation to improve its scalability for industrial-sized instances. We then introduce two new solution approaches: a Logic-Based Benders Decomposition (LBBD) and a matheuristic algorithm, both tailored for medium to large problem sizes. In addition, we develop a tabu search metaheuristic to the FJS-OT problem, building on a version known for its effectiveness in handling industrial-scale flexible job shop instances. This collection of approaches offers scalable decision support, depending on the problem size, available computational resources, and optimality requirements. Computational experiments demonstrate that LBBD outperforms the MIP model on small- to medium-sized instances, while the matheuristic efficiently delivers high-quality solutions for larger cases. For very large instances, the tabu search method proves capable of rapidly generating feasible solutions where other methods may fall short. This tool is currently undergoing testing in industrial environments, providing practical guidance for production planners facing challenging FJS-OT instances.