Manufacturing Real-time Process Plan Selection Based on System Performance

Today’s fast‐ pace manufacturing company environment requires that manufacturing decisions be made at the time when their impact will be the most relevant for the operations. The importance of the proper information and knowledge necessary for such decision can not be overstated. This paper discusses a procedure for the selection of process plans in real‐ time, during the manufacturing execution. Necessary data and knowledge for successful decision are elaborated. Process planning procedure is preprocessed for the knowledge intensive component using rule-based system, but the final decision related to machine and tool assignment has been postponed for manufacturing in real time. Rule-based system is developed using expert systems shell tools with knowledge related to manufacturing features, machining processes and process capabilities. The result of preprocessing is a process plan network of alternatives plans. Real time selection consults the process plan network and selects next processing task (feature). Three step procedure for selection of the feature, its process parameters, machines, and tool s based on the system performance in real time (due dates, the current cycle times, utilization, and WIP) is explained. The feature is selected based on feature precedence network (FPN). Process, machine and tool are selected after ranking the current machine utilization or WIP of all candidate resources and selected the least loaded resource. The procedure is repeated at predefined decision points in order to increase output and balance resource utilization. The procedure has been developed in context of FMS supported by data collection about the system status (utilization and WIP). The procedure is supported by few illustrative examples of simulated FMS system.