Enhancing Last-Mile Delivery Efficiency through Crowdsourced Workforce Scheduling and Routing Optimization

Abstract

This thesis investigates the optimization of last-mile delivery by leveraging crowdsourced workforce scheduling and routing. The study introduces a novel integration of crowdsourcing with traditional workforce scheduling, emphasizing collaborative tasks that involve both in-house drivers and crowd drivers. The developed Mixed-Integer Linear Programming (MILP) model aims to minimize total delivery costs while meeting service level targets, incorporating variables such as worker preferences, time windows, and parcel sizes. Utilizing real-world floating population data, the model dynamically adjusts crowd driver compensation based on population density, ensuring cost efficiency in densely populated areas. Through five detailed case studies conducted in various districts of Busan- jingu, the research demonstrates substantial economic benefits, with crowdsourcing reducing mean delivery costs by 22.61% and median costs by 23.72%. Sensitivity analysis identifies three key insights: the impact of variable costs for in-house and crowd drivers, the influence of time window widths on total costs, and the effects of varying crowd driver compensation. These findings underscore the potential of crowdsourcing to enhance efficiency and reduce costs in last-mile logistics, providing valuable insights and practical solutions to meet the growing demands of the e-commerce sector.