Inventory Management for Integrated Circuit Packaging Material under Variable Demand

Abstract

This paper proposes an appropriate ordering policy to improve the efficiency of inventory management for the packaging group of 24 items in a semiconductor factory, this is to reduce total cost while maintaining service level at 95%. The total cost is comprised of ordering cost and inventory holding cost. This study was conducted as follows. Firstly, all relevant data (e.g., inventory data, ordering conditions, and constraints) is reviewed. Then, the demand is divided into 2 patterns: 1. Smooth demand 2. Erratic demand. So, the grouped demand will be defined by the inventory policies. The concept of policy design is to be replenished at any time because the inventory is continuously reviewed. Thus, the Re-Order point (ROP) policy and (s, S) policy are proposed for Smooth demand pattern and Erratic demand pattern consecutively.

The proposed policies are tested with demand data from 2019 and 2020. The results show a total reduction from 111,832 USD to 46,288 USD in 2019 or decreased by 58.61%, and from 132,886 USD to 51,404 in 2020 or decreased by 64.29%. While the policy is still able to maintain the service level as Fill Rate in 2019 and 2020, the average per year is 95.13% and 95.36% respectively. Additionally, the sensitivity analysis result shows that the proposed policy is robust to changes in demand patterns. There are only about 13% changes in total cost compared to 50% changes in average demand or the worst-case scenario for demand changes.