Optimal supply chain replenishment under uncertain future demand

Abstract

In this research the optimal replenishment for a simple, capacity-constrained supply chain is determined when a demand rate probability distribution is known. Continuous ­review reorder point replenishment is considered. Optimization of decision variables is based on an economic measure, with a profit maximization objective that incorporates holding, tardiness and lost sale costs. A key contribution is the development of a sampling methodology that allows simulation-optimization to be used for this problem. Results using a demand rate probability distribution for optimization are compared with those using only the expected demand rate. Higher profits are observed when using the probability distribution. This research also studies the performance robustness to forecasting error. A new robustness index is developed to evaluate profit sensitivity to deviations from the expected demand. Findings reveal performance is less sensitive to deviations between the actual and expected demand when decision variables are optimized using the demand rate probability distribution.