O2O on-demand delivery optimization with mixed driver forces

Abstract O2O (Online to Offline) service enables customers to place orders online and receive products/services offline. In addition to traditional in-house drivers, the emergence of crowd-sourced drivers provides an opportunity to re-organize the offline delivery. In practice, three types of workforce, namely, in-house drivers, full-time and part-time crowd-sourced drivers, coexist in the system with different characteristics. This posts challenges to the management of order assignment and routing by the online platform. In particular, we study the impact of the detour flexibility for part-time crowd-sourced drivers, which affects the participation rate of this type of workforce. This paper aims to provide a systematic method for the O2O platforms to optimize order assignment and routing. We further validate our model and study the managerial insights using real datasets from one of the mainstream O2O platforms in China.