Development of inventory policies for managing cosmetic ingredients: a case study of a chemical distributor in Thailand

This Independent Study, conducted as part of the SE799 course in the Logistics and Supply Chain System Engineering Program, focuses on optimizing inventory control policies for two cosmetic ingredients, one with high turnover and another with average demand under uncertain conditions. Using 35 weeks of real weekly sales data, empirical distribution functions (EDFs) were developed to represent actual demand variability without assuming a predefined statistical distribution. A periodic review (T, s, S) inventory policy was applied, accounting for real-world operational constraints such as minimum order quantity (MOQ), ordering cost, holding cost, and express delivery in case of shortages. To improve customer service and reduce lost sales, a shortage dissatisfaction cost was added to every shortage event, in addition to the express shipment cost. Two optimization techniques, grid search and golden section search were used to identify the most cost-effective (s, S) values. For the high-turnover ingredient, optimized policy parameters successfully minimized the total average cost while satisfying demand more reliably. For the average-demand ingredient, the results confirmed that holding inventory was not economical, and an on-demand ordering strategy was more appropriate. This study highlights the value of combining empirical data, simulation, and customer service considerations in inventory planning. The findings offer practical guidance for managing imported ingredients with unpredictable demand and varying service expectations.