AI Summary & Key Takeaways
AI InsightThe data set of the real-life problem that was used was applied to those three methods, and 0.45%, 0.81%, and 1.39% improvements were achieved using the methods, respectively.
📌 3 Key Takeaways
Abstract
The vehicle routing problem (VRP) is of great importance for feed factories that do not work with the dealership system. This is especially important in the Central Anatolian region, where customers’ number of animals is low. Data used in the study came from the order data of a feed mill which operates in Turkey. Before selecting the most suitable VRP software vendor, the logistics manager of the plant was urged to analyse the results with the scope of percent fleet capacity used, service level (on-time deliveries), and total transportation cost incurred. As a requirement of the enterprise strategy, a multi-day planning algorithm was developed to level the daily production capacity of the factory while maintaining minimum transportation costs and maximum service level. It has been determined that better results are achieved with the developed multi-day planning algorithm for both methods of Simulated Annealing (SA), Genetic Algorithm (GA), and our Adapted Large Neighbourhood Search (ALNS) heuristic. The data set of the real-life problem that was used was applied to those three methods, and 0.45%, 0.81%, and 1.39% improvements were achieved using the methods, respectively.