Efficient Order Dispatching and Route Planning for On-demand Delivery

On-demand delivery has surged in popularity owing to its capacity to swiftly cater to customers’ immediate requirements and deliver a seamless and efficient delivery experience. However, the potential for improving delivery efficiency through the operations of order handoff and relay has not been considered in the existing related work. To solve this problem, in this paper, we propose an efficient order dispatching and route planning algorithm (E-ODRP) to improve the delivery efficiency for the system profit maximization. Specifically, E-ODRP consists of the online scheduling and the offline preparation algorithm. In the online scheduling algorithm, first, the load-aware greedy algorithm assigns orders to couriers to minimize delivery costs while ensuring that the order loads among couriers does not exceed a threshold value. Second, the proposed fuzzy logic-based path planning algorithm integrates distance, density, deadline metrics of delivery points for the effective scheduling of courier delivery paths. Third, the handoff triggering condition determines the handover area by the discretization method to improve delivery efficiency. In the offline preparation algorithm, the modified K-Means++ algorithm clusters the locations of the historical orders and set the centers of the clusters as the locations of the relay points. The proposed estimated revenue determines the number of the established relay points. We conduct extensive experiments based on real-world data from an on-demand delivery platform in Chengdu, the results show that our E-ODRP outperforms the state-of-art approaches in terms of overall system profit and total driving distance.