Home // SIMUL 2023, The Fifteenth International Conference on Advances in System Simulation // View article

Crowdshipping with Dynamic Workers Availability – Restless-Bandit-Based Prioritization

Authors:
Amin Karimi
Jing Fu
Lele Zhang
Hadi Ghaderi

Keywords: stochastic process; restless bandits; crowdsourcing; task assignment; crowdshipping; parcel delivery

Abstract:
Exponential growth for last mile delivery demand has created several challenges for retailers and couriers, at the same time forcing the development of efficient and sustainable delivery solutions. One of the emerging solutions is crowd-sourced delivery, also known as crowdshipping. In a crowdshipping system, the general public participates in parcel delivery (known as crowdshippers) and then rewarded with remunerations. To develop sustainable and commercially viable crowdshipping solutions, capable of handling large-scale delivery tasks, effective assignment of tasks to crowdshippers is critical. Particularly when both tasks and crowdshippers dynamically arrive and depart the system, it becomes challenging to complete deliveries, while maximizing the total profit of the platform. This paper models the dynamic crowdshipping system using a Markov decision process and proposes a restless-bandit-based capacity relaxation technique to facilitate the task-to-crowdshipper assignment. Simulation results show that the proposed technique is superior over two baseline policies with respect to higher average profits and lower task rejection rates. The learning of this research provides important directions for the design and development of crowdshipping systems that are subject to both crowdshipper and task uncertainty.

Pages: 25 to 33

Copyright: Copyright (c) IARIA, 2023

Publication date: November 13, 2023

Published in: conference

ISSN: 2308-4537

ISBN: 978-1-68558-102-2

Location: Valencia, Spain

Dates: from November 13, 2023 to November 17, 2023