Thesis/Capstone
Publication Date
Advisor(s): Nima Kazemi
Topic(s) Covered:
  • Optimization
  • Forecasting
  • Fulfillment
  • Inventory
Abstract

Spare parts management is an essential operation in the supply chain of many companies, owing to its strategic importance in supporting equipment availability and continuity of operations. In many supply chains, the demand of spare parts is inherently more uncertain compared to traditional fast-moving products. This is due to the fact that spare parts demand is highly intermittent, mostly observed with a long period between consecutive orders, where a no demand period is followed by a period of an order signal. As spare parts are critical to the continuity of operations, companies tend to stock more inventories to mitigate the risk of irregular demand pattern. Gerber Technology, a manufacturing company that sells industrial machines and the spare parts that support them, faces challenges in its spare parts demand forecast quality and inventory management. This challenge has recently been negatively impacting the company’s inventory costs and customer service level, where the actual inventory is consistently higher than the targeted level. Meanwhile, higher inventory levels are not being translated into higher service level to its customers. In summary, the company has seen increased costs with a lower service level. Therefore, the aim of this project was to improve the demand forecast accuracy and the spare parts service level of the company while optimizing inventory costs. For this purpose, we used SKU classification for demand categorization and inventory control. With these categorizations, we then allocate the recommended demand forecasting techniques and optimize the inventory levels of the company. By following these processes, we achieved an improvement between 7% to 14% in forecasting accuracy measured by the Root Mean Squared Error (RMSE). We could also gain up to 3% improvement in service level leading to $1.3M additional revenue opportunity.