| Abstract: | Material requirements planning (MRP) is a planning and information system that has widespread application in discrete-parts manufacturing. The purpose of this article is to introduce ideas that can improve the flow of material through complex manufacturing systems operating under MRP, and that can increase the applicability of MRP within diverse manufacturing environments.MRP models the flow of material by assuming that items flow from work station to work station in the same batches that are used in production. That is, once work starts on a batch of a certain item at a certain work station, the entire batch will be produced before any part of the batch will be transported to the next work station on its routing plan. Clearly, efficiency can be increased if some parallelism can be introduced. The form of parallelism investigated here is overlapping operations.Overlapping operations occurs when the transportation of partial batches to a downstream work station is allowed while work proceeds to complete the batch at the upstream work station. The potential efficiencies to be gained are the following: • Reduced work-in-process inventory • Reduced floor space requirements • Reduced size of transfer vehicles Additional costs may accrue through additional cost of transportation of partial batches and through additional costs of control.Some MRP software vendors provide the data processing capability for overlapping operations. However, the user is given little or no guidance on overlapping percentages or amounts. It is our intent to provide a simple, robust technique to MRP users who would like to overlap operations and gain some or all of the above efficiencies.An optimal lot-sizing technique is derived by considering a generic two work station segment of a manufacturing system. Under the assumptions of constant demand and identical production rates, a cost function that considers setup costs, inventory holding costs and transportation costs is derived. This cost function is minimized subject to the constraint that the production batch is an integer multiple of the transfer batch. We solve for the optimal production batch, the optimal transfer batch, and the integer number relating them. Solutions are obtained as closed form, easy to-evaluate formulas.By introducing more parallelism, overlapping operations can reduce lead time. However, this will not happen without modification of MRP logic to accommodate such reduced lead time. We derive a formula that shows how a significant lead time compression can easily be obtained and implemented in MRP.We consider an example to illustrate the application of the technique on typical data from the electronics industry. The outcome showed a cost savings of approximately 22.5% over the standard MRP approach.Overlapping operations allows the applicability of MRP to an increasing number of situations that are not modeled faithfully by conventional MRP logic. Three such situations that occur often are the following: • Limited size of transfer vehicles dictate that several transfers should be planned. • Lead time requirements prohibit nonoverlapped operations. Our analysis suggests how to accommodate these difficult practical situations into MRP.Overlapping operations in material requirements planning provides an enhancement that allows wider applicability, shortened lead times, and lower total costs. It may be applied selectively to any two work stations where it is deemed appropriate. Due to the structure of the cost function, it is possible to make the transfer lot-sizing decisions independent of the production lot-sizing decisions. Therefore, significant improvements can be made through overlapping with minimum disruption to the existing MRP system machinery. It is our conviction that overlapping operations is an important concept that can and will impact MRP. We suggest the approach presented here as a systematic way to implement overlapping. |
