A multiobjective evaluation of flexible manufacturing system loading heuristics

Modern Flexible Manufacturing Systems (FMSs) are implemented to accomplish highly efficient, automated, concurrent production of several part types. The loading problem is the portion of the short to medium term FMS planning problem concerned with allocating operations and tools to machines subject to machine time and tool slot capacity restrictions. It is assumed that the set of parts along with their production goals for the period are specified. Heuristics can be readily constructed for specific objectives, but the real world problem has multiple objectives. In this paper the authors compare the performance of six loading heuristics on three existing FMSs. The heuristics differ in the objectives explicitly considered and in the relative priority assigned to the objectives which are considered. Performance of the heuristics is based on separate evaluation of the five objectives: workload balance, volume of intermachine part movement, routing flexibility, tool investment and maximum machine utilization. In general, performance on the criteria is seen to differ in accord with the heuristics’ objectives. However, assigning operations to machine types based on workload balance performs better than assigning to the most efficient machine type.