Modeling, validation and control of manufacturing systems

In this paper we elaborate on the problem of supply chain control in semiconductor manufacturing. First, we introduce the problem. Next, we propose the use of effective processing times (which can be measured from real factory data) to arrive at 'simple' discrete-event models for manufacturing systems. We explain why existing models can not be used for solving the problem and explain the need for PDE-models that consider the flow of products as a compressible fluid flow. Next, we present a validation study in which we compare the response of the currently available PDE-models to the results of discrete-event simulation. We conclude the paper by analytically deriving a controller that solves the ramp-up problem. The resulting controller is often used in practice.