Abstract
This paper first formulates a binary integer programming (BIP) model that minimises C max for a multi-stage flexible job shop floor with machine compatibility. Due to a computational limitation, the exact optimal model is then relaxed as a linear programming model. The output from the relaxation model then turns into the objective of the BIP-based real-time scheduling (RTS) heuristic model. The RTS heuristic requires an iteration to calculate the final C max. At each iteration, the RTS heuristic assigns just one job to the earliest available machines. Since the set of jobs and machines included in the RTS model is relatively small, RTS can be solved in a very short computational time. We evaluate an overall effectiveness (in terms of solution quality and run time) of the RTS heuristic by way of computer experiments.