Multi-warehouse, multi-product inventory control model for agri-fresh products – A case study

Abstract

Warehouses that operate under controlled micro-climates are necessary to reduce the deterioration of fresh produce, but they are also costly to operate. Developing an efficient operation regime for these premises while considering energy costs, demand, and harvesting requirements is crucial to managing the food supply chain sustainably and profitably. This paper proposes a method to develop an operation regime by solving a multi-period, multi-product, multi-warehouse inventory control optimization problem that originates from the management of agri-fresh products from an Australian company. We propose a mixed-integer quadratic programming model for this problem. A significant challenge is that each product has a specific deterioration rate over storage, which differs in each warehouse mode. Our model allows for two different operational modes for warehouses, each with different electricity consumption, affecting the time the products remain fresh. The objective function of the model minimizes the total inventory costs. The model provides the optimal inventory flow of each product, the optimal number of active warehouses, the optimal operational length of warehouses, and the optimal mode of warehouses in each period. We solve the model using Gurobi and integrate three other solution approaches based on model properties, which reduce the computational time by half. The model is tested using real data from the case study, an Australian agri-fresh company. The results indicate a reduction in total cost by 8% and quantity of product deterioration by 20%. We also investigate the impacts of variations in demand and supply caused by growing or shrinking markets in our case study under different scenarios and instance sizes of the problem and analyze the impact in solution times that are observed in practice.