Modelling short-term manufacturing flexibility by human intervention and its impact on performance

Short-term manufacturing flexibility is the capability to respond to unforeseen short-term and immediate events during operation. Explicitly modelling short-term manufacturing flexibility is extremely hard as it implies that it is possible to prescribe responses to all possible events in all possible states of the system under consideration. We propose an implicit modelling approach where we only model the frequency of responses to unforeseen events and the impact of the response on the evolution of the process under consideration. This allows us to dramatically reduce model complexity while preserving empirical validity of the model. We introduce the concepts of intervention–independent performance (IIP) and intervention–dependent performance (IDP). We present a methodology to gather IIP and IDP indicator information. We propose another process to use this information together with historical transactional data and forecasts to make tactical trade-offs that result in control policy parameters. We illustrate our modelling approach by a generic inventory control model with replanning of replenishment orders. We apply the model in a case study characterised by highly utilised production lines, high set-up times and dynamic demand. We provide empirical evidence of the validity of the proposed methodologies.