When there is a production system with excess capacity, i.e. more capacity than the demand for the foreseeable future, upper management might consider utilizing only a portion of the available capacity by decreasing the number of workers or halting production on some of the machines/production lines, etc. while preserving the flexibility of the production system to satisfy demand spikes. To achieve this flexibility, upper management might be willing to attain some pre-determined/desired performance values in a production system having identical parallel machines in each work center. In this study, we propose a framework that utilizes parallel neural networks to make decisions on the availability of resources, due date assignments for incoming orders, and dispatching rules for scheduling. This framework is applied to a flexible manufacturing system with work centers having parallel identical machines. The artificial neural networks were able to satisfactorily capture the underlying relationship between the design and control parameters of a manufacturing system and the resulting performance targets.

This is the author's version of the work. It is posted here by permission of Elsevier for personal use, not for redistribution. Accepted for publications to Computers and Industrial Engineering,Vol. 50, No. 1, May 2006 doi:10.1016/j.cie.2006.02.002