PERFORMANCE ANALYSIS OF FLEXIBLE MANUFACTURING SYSTEMS WITH A SINGLE DISCRETE MATERIAL-HANDLING DEVICE

We present an analytical model for performance prediction of flexible manufacturing systems (FMSs) with a single discrete material-handling device (MHD). This configuration of FMS is significant for many reason s: it is commonly found in industry, it simplifies material-handling c ontrol, it is amenable to analytical modelling, and it forms a buildin g block for more complex systems. Standard queueing models are inadequ ate to analyze this configuration because of the need to take into con sideration many nontrivial issues such as state-dependent routing, int erference from the MHD, and the analysis of the MHD. To account for st ate-dependent routing, we develop an iterative method that is built ar ound mean value analysis. To analyze the MHD interference, we use two queueing network models. In the first, we ignore queueing at the MHD b ut model the interference from the MHD by inflating the station servic e times. The second network models the queueing for the MHD and estima tes the blocking (inflation) times needed for the first model. By iter ating between the two networks, we are able to predict the performance of this configuration of FMS. Our analytical estimates are validated against discrete event simulation and shown to be quite accurate for i nitial system design.