Data-Driven Scalable E-commerce Transportation Network Design with Unknown Flow Response

Abstract

Problem definition: Motivated by our experience with a large online marketplace, we study an e-commerce middle-mile transportation network design problem. A salient feature in this problem is decentralized decision-making. While the middle-mile manager decides the network configuration on a weekly or bi-weekly basis, the real-time flows of millions of packages on any given network configuration (which we call the flow response) are controlled by a fulfillment policy employed by a different decision entity. Thus, we face a fixed-cost network design problem with unknown flow response. Methodology/results: To meet this challenge, we first develop a predictive model for the unknown response leveraging machine learning techniques. Apart from the most natural network-level predictive model, we find that the more parsimonious destination-level and arc-level predictive models are more effective. We then embed the predictive model into the original network design problem and characterize this transformed problem as a c-supermodular minimization problem. We develop an approximation algorithm whose running time scales linearly with the number of potential arcs and has additive error depending on the constant parameter c. We demonstrate that this algorithm is scalable and effective in a numerical study. Managerial implications: We introduce a novel transportation network design problem by proposing the notion of unknown flow response, particularly relevant in e-commerce fulfillment process involving granular shipments and long-tail demand. Our analytical approach offers e-retailers a more efficient way to optimize their networks compared to costly simulations. Our method is also applicable to middle-mile network design for omnichannel retailing.

Keywords: E-Commerce Fulfillment, Middle-Mile Transportation, Machine Learning, Approximation Algorithm