A study of bottlenecks in a MRP environment using simulation

In this paper a simulation study is described which investigates the manufacturing operations of an engineering company. Thirteen primary end items are assembled from purchased items and components fabricated in their machine shop. The production scheduling function is based on a computerised Material Requirements Planning (MRP) system which generates component and assembly schedules from a Master Production Schedule (MPS), the bills of materials for the products and the component inventory file. These schedules are in constant jeopardy because of invalid time data and delays causing late deliveries. From a preliminary analysis it was found that a serious bottleneck existed on one machine in a system of eight machine groups. A simulation of the actual operations was performed which verified the existence of the bottleneck. Sensitivity testing revealed the dramatic effect on total system performance of increasing the capacity of the bottleneck machine. When the bottleneck was eliminated, total throughput was increased, inventories reduced, and the accuracy and validity of the MRP schedules were significantly improved. These improvements were achieved in significant part due to reductions in both the mean and variance of the manufacturing lead times.