An operational policy for a single vendor multi buyer integrated inventory supply chain system considering shipping time

Since its introduction, the concept of integrated inventory supply chain has received a considerable amount of attention. The majority of studies in the last three decades revealed an increase in holding cost as product moves further down the chain or up the chain. A recent study Hoque (2008) considered vendor’s setup cost and inventory holding cost. Some research also considered fixed transportation cost, which is unrealistic. This study focuses on a single-vendor, multi-buyer scenario and presents three models. First, two models illustrate the transferring of equally-sized batches. Then, a third model considers the transferring of unequally-sized batches in a lot. This study relaxes the assumption that vendor’s holding cost must be greater than or less than all buyer’s holding costs in the system. Also, this research facilitates unequal transportation time and cost for different buyers for greater flexibility. The total system cost is calculated by summing the annual operational cost for all the parties in the system. Optimum values of the decision variables are determined using a direct search method. As presented by the third model, a numerical example demonstrates that the total system cost is less when compared with other two models presented. This study also presents the following: solution procedures to solve each model, many numerical examples to support mathematical findings, and performance comparisons among three findings. In order to justify the lot-splitting approach for solving the integrated inventory problem, alternative models with no lot splitting are devised and tested under the same circumstances. Alternative models with no lot splitting produce similar or better results. Under the same circumstances, the alternate third model is observed to be offering the least total cost for the system. This study also presents a sensitivity analysis to check the robustness of the three models. The future extension of this research may involve considering storage capacity constraint and random demand