Evaluating a Crowd Logistics Network Using Agent-Based Modeling

Crowd logistics is a part of the sharing economy in which individual carriers offer to transport and deliver items for other individuals or businesses for a fee via an online platform. While crowd logistics platforms have the potential to offer more flexible and responsive delivery services for much lower rates than traditional logistics providers, it is challenging for platforms to achieve a critical mass of participants on both sides (senders and carriers) to allow the service to grow and thrive. This research uses agent-based modeling to explore the effects of participant behavior on the performance of a two-sided crowd-sourced logistics platform. Preliminary experimentation with the model tests the effects of heterogeneous agent decision logic on platform performance, including service level and network growth over time. Results demonstrate significant differences in performance between heterogeneous and homogeneous decision rule assignment and suggest that agent-based modeling is a particularly suitable method for studying the behavior of crowdsourced platforms.