Manufacturing system design with optimal diagnosability

Abstract

This paper addresses the quality issues of advanced manufacturing systems that require frequent reconfiguration to satisfy customers' demands. The approach embodies the diagnosability issues in the design of such systems so that monitoring and identification of the symptoms of the critical quality characteristics can be carried out effectively. A diagnosability matrix model is constructed to delineate the relationships between attributes of quality monitoring and fault diagnosis. Based on this matrix, the conditions of optimal diagnosability are established. Then, a systems approach which is based on axiomatic design theory is applied to achieve optimal design for diagnosability (DFD). An industrial case study demonstrating the application of the optimal DFD approach is described.