A comparison of schedules resulting from priority rules and mathematical optimization for a real production cell
Conference contribution, 2010
In this article we present a complete algorithm using mathematical optimization tools for solving a flexible job shop scheduling problem at Volvo Aero, a company producing high-valued low volume components for the aircraft industry. The goal of the scheduling is to facilitate a higher utilization of the cell while minimizing the total tardiness and the cell throughput time. The production cell consists of ten resources (machines and workstations) whereof five are multipurpose machines that can carry out a variety of operations; the so-called multitask machines work in parallel instead of one dedicated machine for each product. The production cell is studied from a mathematical as well as a logistical point of view. The quality of the schedules is measured by means of the total tardiness and computation time. The resulting schedules from a mathematical optimization model are compared with schedules generated by priority rules, which are similar to today’s manual scheduling of the multitask cell. The tests were carried out for different realistic scenarios with regard to work load and product mix. The resulting schedules, which as expected outperformed the two commonly used priority rules, are obtained within minutes for the upcoming shift.