MRP-controlled batch-manufacturing environment under uncertainty

The overall aim of this research is to model the effects of uncertainty on delivery performance in an MRP-controlled batch-manufacturing environment with multi-product and multi-level dependent demand. To this end, MRP planning and batch-manufacturing system control architectures were modelled using simulation to implement the MRP release logic. Simulation and experimental design were carried out based on a real case enterprise. ANOVA showed that four uncertainty factors—namely late delivery from suppliers, machine breakdowns, process batch size increments and customer design changes—have significant effects on delivery performance. This ANOVA further showed that uncertainties create knock-on and compound effects; the latter are difficult to predict in practice. Significant two-way and three-way interactions among some uncertainty factors were also found, making it more difficult to characterise the precise factor effects. It was found that the more uncertain the environment is, the later the deliveries are. It can be concluded that MRP-controlled batch-manufacturing enterprises should diagnose uncertainties that are significantly affecting delivery performance, and tackle these uncertainties most urgently to prevent diffusion of knock-on and compound effects and improve delivery performance. This conclusion was validated through the case enterprise, for which significant delivery improvement has been achieved.