Production Variability And Its Approximate Technique For Manufacturing Systems

Recent remarkable advances in manufacturing systems, such as agilemanufacturing, have regarded production variability as an important designcriterion, whose purpose ensures both predictability and dependability ofmanufacturing systems.This paper develops an approach for analyzing the production variabilityof manufacturing systems. Two problems are addressed: the problem of theoutput variance and the problem of the delivery-time variance. This approachis based on approximating the manufacturing system as a Markovian ArrivalProcess (MAP). By using the results developed for the MAP, the productionvariability of the manufacturing system can be determined.Three kinds of manufacturing systems are analyzed: serial productionline with reliable machines, serial production line with unreliable machinesand unreliable module-based assembly (MOBA) system. Numericalexperiments show that this approach has many advantages in the study ofproduction variability of complicated manufacturing systems which havebeen an important topic in this area up to now. In this approach, the statenumber is reduced through MAP approximation subsystem by subsystem, sothe total computation effort for production variability is controlled within areasonable range. Relationship between the system parameters and theproduction variability are also investigated by numerical experiments.It is believed that this study provides a basis for a promising approachfor production variability evaluation in complex manufacturing systems.