A spatial agent-based model of e-commerce last-mile logistics towards a delivery-oriented development

The spread of e-commerce has driven a major growth in the parcel delivery market, bringing a negative impact on sustainability, especially due to last-mile deliveries in urban areas. It is crucial to appropriately tackle this issue to foster consolidation of deliveries, possibly by using collection and delivery points (CDPs), where customers might receive their parcels. This paper proposes a new spatial agent-based modelling approach to explore different scenarios of last-mile logistics referred to e-commerce deliveries, comparing fragmented door-to-door deliveries with consolidation-based strategies. The case study is the central urban area of Catania, a medium sized city in Southern Italy. The Agent-Based Model (ABM) reproduces feasible operations considering real-world spatial constraints and demand data, including the possible matching of customers' systematic trips and parcel delivery via CDPs with small detours from the scheduled trip. Key performance indicators consider both customer and logistics operator perspectives. Main results of the simulation show that the scenario without CDPs is the costliest and least efficient, implying a high number of failed deliveries. Using cargo bikes instead of vans to perform the delivery implies high costs, but much higher benefits in terms of reduced energy consumption. The highest logistics efficiency is achieved in the scenario with a doubled demand, implying a better use of the CDPs. The results suggest that it is advisable to incentive the use of CDPs instead of increasing their number. The ABM can provide useful information to decision-makers on how to manage growing on-demand urban deliveries and plan last-mile logistics using a delivery-oriented development approach.