Use of GI/G/1 queuing approximations to set tactical parameters for the simulation of MRP systems

There is a lack of prescriptive methods for setting lot sizes and planned lead times effectively in MRP systems. Recent research has suggested the application of queuing relationships. This study experimentally investigates the use of GI/G/1 relationships for lot size selection along with the use of exponentially smoothed feedback for dynamic planned lead time setting. Results show that assumptions regarding lot interarrival time variability have a large effect on lot sizes and performance.