INVENTORY ROUTING PROBLEM IN ROLLING HORIZON PLANNING ENVIRONMENT

This article develops a solution approach for the Inventory Routing Problem in rolling horizon planning environment. The authors try to identify potential advantages of vendor managed inventory (VMI) approach implementation for consolidated deliveries. The main goals of the paper are identification and evaluation of potential savings due to replacement of classical vehicle routing problem (VRP) with the inventory routing problem (IRP) approach. We compare the distribution plans based on classical VRP algorithms to IRP-approach-based plans for different length of planning horizon. Our purpose is to create a solution methodology that could be applied for real-life instances that are critical for company's resource planning and fleet utilization. In order to solve the test instances, two complementary approaches are taken into consideration. The first one is an extension of exact model for solving routing problems given by Miller et al. (1960). Due to the complexity issues (only test instances up to 6 locations could be solved to optimality) heuristics are also introduced. Our heuristics proposal allows obtaining close to optimal solutions using mixed integer programming (MIP) and is applied to find a solution for larger instances. Computational results of experiments, demonstrating the effectiveness of our approach, are reported.