Optimizing Parcel Locker Locations in a City Crowd Logistics Network

Abstract

There is a growing market for crowd-shipping (CS), which hires people to transport parcels on their regular commutes. This paper considers a hybrid CS operation that hires crowd-shippers and regular drivers (RDs) in a physical internet environment. A city crowd logistics (CCL) network is designed to serve business-to-consumer logistics demand through solving a mixed-integer linear program of a parcel locker location assignment problem. The problem minimizes the total cost by optimizing the placement of parcel lockers and the movements of parcels. As an input to the program, the spatial distribution of potential crowd-shippers is analyzed by leveraging a behavior model and data that combines household travel survey data from the Greater Toronto Area (GTA), Canada, and Canadian census data. Parcel demand is simulated based on the proportions of retail employment and the population of the corresponding planning district in the GTA. The numerical results reveal that optimal parcel locker locations are more likely to be placed in Toronto while less likely to be installed in Durham and Halton regions. Concerning the service levels of the proposed network, it is found that a CCL network can provide more stable and cost-effective services when the parcel demand level is higher. The hiring of RDs is necessary to service some regions with low crowd-shipper supply. Insights from this study can help logistics companies to develop an effective business strategy for a hybrid CS operation system.