Designing Optimal Hub-and-Spoke Network Considering Flow Delay Cost: The Case of Urban Express Delivery in Eastern China

Abstract With respect to a traditional point-to-point (P-P) network, a hub-and-spoke (H-S) not only uses a smaller number of links/paths, but also utilizes the scale economy advantage on consolidated flows on hub-hub links and at hubs. However, the inevitable delays through hubs have always been a critical concern. Therefore, this paper develops an H-S model considering flow delay costs and applies the model to a logistics case in east Jiangsu province, China. The model is based on the well-known H-S network model proposed by O'Kelly et al., which first introduced the concept of flow delay cost in hub location and network design. In this research, the flow delay cost is defined as the cost associated with delay caused by flow transfers at hubs. The integer quadratic term in our model’s objective function is linearized using the algebraic method suggested by Shen (1994). Our model is applied to a local logistics firm, who plans to develop an H-S network for its 13-node express package delivery operation, using the Particle Swarm Optimization (PSO) algorithm. The results show using the H-S can save more than 14.1% of the total cost annually. The model also provides an applied case to the H-S configuration, especially for urban express delivery logistics in China.