Scheduling heterogeneous delivery tasks on a mixed logistics platform

Abstract

Large e-commerce retailers usually establish their own logistics systems. Such systems make use of their own dedicated fleets but will also use a crowdsourced delivery mode by hiring occasional fleets. These mixed logistics systems with both dedicated and occasional fleets serve both retailers’ internal delivery tasks and external tasks requested by local businesses. This paper studies the problem of scheduling heterogeneous (internal and external) delivery tasks on a mixed logistics platform with multiple depots and two types of vehicles (dedicated and occasional). A delivery task is executed by either a dedicated vehicle or an occasional vehicle. The dedicated vehicles depart from and return to the platform's depots; the occasional vehicles depart from their original location and pick up goods from depots or external pickup locations, fulfill the delivery tasks, and finish their route at the final delivery location. We propose mixed integer programming models and column generation-based solution methods to solve the problem. A computational study is conducted based on a series of randomly generated instances and real-world instances involving 15 depots, 120 internal customers, 15 external delivery tasks, and 38 dedicated and occasional vehicles. The results obtained demonstrate the efficiency of the column generation-based solution methods. Moreover, the effectiveness of the proposed models is validated by a significant cost saving in comparison to intuitive decision rules. A sensitivity analysis is also conducted to derive a number of managerial implications.