Though the issue of vehicle routing in a distribution center has received immense attention from researchers in the field of material handling, the focus, in a majority of cases, has always been towards large scale AS/RS Systems and AGV systems in a layout devoid of intermediate travel aisles. This thesis work measures order processing times and the associated vehicle congestion in small to medium scale distribution centers where the order picking activities are performed by manually operated forklift vehicles. This is accomplished by designing an experiment where the prominent factors are item storage policies, vehicle routing strategies and warehouse dimensions. In essence, the ABC storage and the random storage policies are crossed with four routing strategies which are, in turn, crossed with 81 different warehouse layouts thus giving rise to 648 individual combinations. Algorithms are written in MATLAB to generate orders for the two storage polices and the corresponding routings. The 81 warehouse models are created in the Pro-Model simulation software. By simulating the order picking and vehicle routing events for each combination, the time to process orders and the amount of congestion in the system are measured and reported. The most interesting finding of this thesis is that the routing strategies seem to have an effect on congestion when the items are located based on the ABC storage policy but not when they are based on Random storage. Moreover, a new routing policy called BOSR policy is proposed which performs better than the K-Shortest Paths policy in terms of blocking.
展开▼
