Modeling Fleet Management for The Urban Distribution of Perishable Goods According to Customer Proximity

Managing the distribution of perishable goods is a major challenge due to the need to maintain their quality throughout the transport process. ‎Fleet management plays a crucial role in this regard. This paper proposes a fleet management model that integrates customer segmentation ‎based on their proximity to the depot. Customers are classified into two categories: nearby customers and distant customers. This is made ‎possible by adding a grouping constant. Ordinary vehicles serve nearby customers, while refrigerated vehicles serve distant customers to preserve product quality by regulating temperature. This distinction allows us to adapt delivery strategies, optimize routes, and guarantee the quality of the products delivered, while reducing logistics costs. This will therefore reduce product waste through strict control of ‎delivery times and reduce the carbon footprint through the rational use of energy-intensive refrigerated vehicles. The grouping constant has a ‎strong influence on the model's performance, which is why a sensitivity analysis was carried out to select the ideal grouping constant for ‎solving the developed model. Experimental tests were carried out on instances from the literature. The results of the experiments show that ‎for the CPLEX solver, 76.17% of instances were solved within the allotted time, compared to 23.83% of instances that were not solved ‎within the allotted time, while the decomposition heuristics solved all the instances used. Comparisons between the CPLEX solver and the ‎decomposition heuristic (TPDH) were made on instances for which we obtained solutions with both methods. The results show that the ‎average cost difference between the two methods is 4.70%. They also show an average gain of 94% in computation time in favor of TPDH‎.