A simulation-optimisation approach for production control strategies in perishable food supply chains

ABSTRACT In order to minimise wastes and losses in perishable Food Supply Chains (FSCs), it is crucially important to control inventory and organise the flow of material and information throughout supply chains. In this study, Basestock-Constant Work-in-Process (B-CONWIP), a pull-based inventory control policy, is proposed to control inventory in perishable FSCs. The effectiveness of the B-CONWIP policy is investigated for a three-echelon perishable FSC using a simulation-optimisation approach. The B-CONWIP policy is compared with two existing basestock policies (BSPs): continuous review (s, S) policy and BSP-low-Estimated Waste (BSP-low-EW), a periodic inventory review policy that outperforms other policies presented in the literature for perishable inventory. The objective is to minimise the total cost (i.e., sum of holding, deterioration, ordering, and shortage costs) while satisfying a predetermined service level. The study shows that B-CONWIP yields the lowest total cost, BSP-low-EW performs the second best, while (s, S) policy is the worst. It is also noted that, in spite of higher ordering cost for B-CONWIP compared to (s, S) policy and BSP-low-EW, B-CONWIP achieves the lowest total cost at all demand rate and volume variations tested. In conclusion, B-CONWIP is more flexible and robust than (s, S) and BSP-low-EW policies because of its ability to handle demand variations without major changes in control parameter values and cost measures.