A matching game-based crowdsourcing framework for last-mile delivery: Ground-vehicles and Unmanned-Aerial Vehicles

Abstract

This paper tackles the problem of allocating time-constrained last-mile delivery tasks to crowdsourced vehicles and Unmanned Aerial Vehicles (UAVs). With the increasing development of E-commerce, the demand for home deliveries is rapidly increasing. However, last-mile delivery remains the costliest part endured by shipping companies, particularly for time-critical deliveries. To mitigate this high cost, recently, several researchers and companies, such as Amazon and Mercedes-Benz, have adopted (1) crowdsourced and (2) UAVs-based delivery solutions. Nevertheless, crowdsourcing delivery tasks, where workers with vehicles are being recruited to deliver parcels, suffers from long delivery time and delays for time-critical tasks during peak traffic hours. Although UAVs are perceived currently as a trending solution for fast and timely delivery of time-critical delivery tasks, yet they are constrained by the flying time and the package weight. These limitations have a direct impact on the cost of delivery and the task accomplishment rate. In this paper, a hybrid framework that crowd-sources vehicles’ drivers and UAVs for delivery tasks is proposed. It aims at improving the delivery time and the allocation percentage mainly at peak hours, by the proposed, (1) task filtering, (2) worker quality estimation, and (3) task allocation and scheduling. The filtering of possible tasks, that each type of worker can accomplish is done based on the availability of the workers and their ability to reach the destination. The worker quality, with regard to the delivery task, is estimated using the Bayesian belief function that is utilized in the allocation. The allocation of tasks and workers is performed using the Gale–Shapley matching game to consider the preferences of both workers and tasks. The simulation results show that the hybrid framework outperforms the non-hybrid crowdsourcing model in terms of task allocation percentage, delivery time, and payoff.