Abstract
A learning approach is presented to the design of comprehensive production control systems for complex situations. The approach essentially starts from the assumption that defining the production control problem and solving the problem are interdependent processes. Furthermore, the approach explicitly takes into account the fact that in complex situations, a priori knowledge about the problem situation is Beverly limited especially with respect to organizational circumstances.
As a test, the approach has been applied to the problem of designing a production control system for a diffusion department of a semi-conductor plant. The different phases of the design process in practice are described and the results of each design phase are presented. A rough survey is given of the growth of the control system together with an account of the performance of the final control system.