System dynamics comparison of three inventory management models in an automotive parts supply chain

Abstract

Background: The automotive parts supply chain measures its success in terms of parts availability and stock required to achieve the availability target, measured as allocation fill rate (AFR). The supply chain strives to achieve an AFR target of 95.5% while maintaining low levels of stock.Objective: The first objective of this study is to evaluate the current inventory management approach, namely the maximum inventory position (MIP) method, to understand the difference between the theoretical derivation and the actual implementation. The second objective is to develop and compare the performance of a new stock target setting (STS) method relative to the MIP methods.Method: The theoretical and actual equations behind the MIP and STS methods are derived for steady state as well as stochastic conditions. A system dynamics simulation model (SDSM) was developed to describe both the local and imported supply chains. The SDSM was used to simulate and confirm the parameters for the STS method. It was also used to compare the three inventory management methods against a theoretical environment and actual data sets.Results: The STS method requires a damping factor (DF) to ensure it does not cause the bullwhip effect. The SDSM was used to determine that a value equal to the lead time ensures effective damping. In the theoretical environment, the MIPTheory method requires the lowest stock, but also has the lowest AFR. MIPActual achieves the highest AFR, but with significantly higher stock holding. The STS method improves on the AFR achieved by the MIPTheory method, with lower stock holding than the MIPActual method. With the actual demand data sets, the results vary by parts movement type. With fast moving parts, all methods achieve the AFR target, the MIPActual method has a higher stock holding for all cases, and the STS method results in reduced stock holding for 7 of 12 cases. With medium moving parts, the MIPActual method improves on the AFR in all 15 cases, but with significantly higher stock. The STS method increases the AFR in 7 of 15 cases and reduces the stockholding in 11 of 15 cases. With slow moving parts, both the MIPActual and STS methods improve the AFR with increased stock holding. The increase in stock holding for the STS method is significantly lower. With erratic moving parts, the MIPActual method improves on the AFR in all 17 cases, but requires significantly higher stock holding. The STS method achieves lower AFR values in 10 of 17 cases, but also requires lower or equal stock holding in 10 of 17 cases.Conclusion: The STS method provides a new approach to inventory management in the automotive supply chain. It provides improved performance for lower stock holding than the implemented MIP method (MIPActual). The results for the different movement category suggest that there is further research to be done to confirm the effectiveness of the various methods with other demand distributions.