Smoothing Short Duration Capacity Problems in Enterprise Resource Planning (ERP) Systems: An Algorithmic Approach

Abstract

Successful implementation of ERP systems should take full advantage of the access to information, but not be constrained by many of the deficiencies associated with infinite capacity scheduling methodologies. In this paper an algorithm is developed which improves the performance of the ERP system. Lot sizing decisions based on capacity availability are used as an instrument to integrate more effectively capacity requirements planning (CRP) and Material Requirements Planning (MRP). MRP is a planning tool for a sub-set of manufacturing system specially in hierarchal multi-product, multi-period and multi-stage production planning and inventory control system. The Vehicle used for this integration is planned order release (POR) quantity. The algorithm requires minimal deviation from the MRP logic. In this sequential process MRP first issues the action notices and then the algorithm analyses the capacity situation in the work centers. The algorithm scans all planned order release quantities of lower level items in the action bucket before orders are released to the shop. For each item a delta value is computed identifying the degree orders are contributing to capacity problems. This delta is specific to each POR. The POR causing the highest overload and the one passing through the lowest utilized work centers are selected for a lot size decrease, respectively, an increase. After identifying the candidates, a series of checks analyses the suitability of the candidates for that change. The planned order release quantities of more items can be performed. Projected capacity profile after iteration will be updated and capacity requirements over a short duration are smoothed. The experiments are performed with MAP/3000 as a simulation model for this study. The developed algorithm is added to MAP/3000 as additional subroutines. The design of the experiments consists of a base test identifying the significance of different environmental parameters of MRP systems and a main test which performs a detailed evaluation. The results are evaluated using analysis of variance techniques. Most ERP systems built on the historical development of MRP and CRP systems, and the assumption of infinite capacity is affecting the performance of those systems badly. This paper is presenting an algorithm to smooth capacity problems by using the existed capacity.