This study is concerned with the application and extension of Wagner-Whitin lot sizing technique to multistage manufacturing planning known as Material Requirement Planning(MRP). A modification of Wagner-Whitin algorithm is attempted to improve its application potential in MRP.
The inherent weakness of Wagner-Whitin algorithm(when applied to MRP framework) is that the inventory level in the last period of the planning horizon is zero. By extending the planning horizon by the lead time of each item, we redefine the planning horizon to which the algorithm is applied, and we can keep the appropriate inventory level in the last period of the original planning horizon. That inventory covers the possible future demands arrising in the forecasting horizon. The future demands in the additional horizon are forecast by known forecasting techniques according to demand patterns.
The another part of this thesis is concerned with the relationship between the lot size and capacity limitation. A heuristic algorithm is developed to levelize the over-capacity lot size down to the feasible region. The criteria of the leveling algorithm is characterized by the setup cost and inventory holding cost. A basic assumption made in developing this algorithm is that the net requirements for each item do not exceed the production capacity of each production level.
To appreciate the results of this study, a case study is performed for a typical automobile engine shop. The total system cost resulted from the application of Wagner-Whitin algorithm is cpmpared with that of "lot for lot". Lastly, the cost increase caused by the capacity limitation is examined.
