Crowdsourced auction-based framework for time-critical and budget-constrained last mile delivery
Esraa Odeh,Shakti Singh,Rabeb Mizouni,Hadi Otrok
DOI: https://doi.org/10.1016/j.ipm.2024.103888
IF: 7.466
2024-10-02
Information Processing & Management
Abstract:This work addresses the problem of Last Mile Delivery (LMD) under time-critical and budget-constrained environments. Given the rapid growth of e-commerce worldwide, LMD has become a primary bottleneck to the efficiency of delivery services due to several factors, including travelling distance, service cost, and delivery time. Existing works mainly target optimizing travelled distance and maximizing gained profit; however, they do not consider time-critical and budget-limited tasks. The deployment of UAVs and the development of crowdsourcing platforms have provided a range of solutions to advance performance in LMD frameworks, as they offer many crowdworkers at varying locations ready to perform tasks instead of having a single point of departure. This work proposes a Hybrid, Crowdsourced, Auction-based LMD (HCA-LMD) framework with a dynamic allocation mechanism for optimized delivery of time-sensitive and budget-limited tasks. The proposed framework allocates tasks to workers as soon as they are submitted, given their urgency level and dropoff location, while considering the price, rating, and location of available workers. This work was compared against two benchmarks to assess the framework's performance in dynamic environments in terms of on-time deliveries, average delay, and profit. Extensive simulation results showed an outstanding performance of the proposed state-of-the-art LMD framework by accomplishing almost 92% on-time deliveries under varying time- and budget-constrained scenarios, outperforming the first benchmark in the on-time allocation rate by fulfiling an additional 24% of the tasks the benchmark failed, with around 50% drop in average delay time and up to x5.8 gained profit when compared against the second benchmark.
computer science, information systems,information science & library science