Comparison of inventory policies for multiple products at warehouse: a simulation study

This study compares the performance of two inventory control policies: the order-up-to-level (s, S) policy and the order-up-to (T, S) policy. The analysis was conducted using a simulation model based on real weekly demand data for three products distributed from a central warehouse to multiple retailers. For the (T, S) policy, two review periods, T = 1 and T = 2 weeks, were tested. Each policy was evaluated using two methods. Method 1 applies Excel Solver with a fill rate constraint to determine inventory parameters, while Method 2 uses a formula without applying the constraint. The simulation covers two years of data. The first year was used for parameter optimization, and the second year for testing performance under demand variability. Performance was measured using average total cost, average fill rate, coefficient of variation (CV), and bullwhip effect (BWE). Results show that (T, S) T = 1 Method 2 achieved the lowest average total cost across both years, making it the most cost-efficient option. In contrast, (T, S) T = 2 Method 1 achieved the highest fill rates, offering more stable service performance. The (s, S) policy delivered consistent results in some cases but experienced the highest cost increase under rising demand variation. When considering product-level performance, (T, S) T = 2 Method 2 provided the best control over order variability and demand fluctuation. These findings highlight the importance of selecting inventory strategies based on both cost efficiency and service level priorities under different demand patterns.