Optimization of two-echelon last-mile delivery via cargo tunnel and a delivery person

Abstract

More and more people live in urban areas in general and in “megacities”, with 10 million inhabitants and more, in particular. One innovative instrument to reduce surface traffic and its negative impact on health, congestion, environment, and safety when supplying these urban inhabitants with goods is a cargo tunnel. Within the cargo tunnel concept, autonomous rail-bound, maglev, or automated guided vehicles deliver goods underground toward small inner-city hubs, from where the final leg toward customer homes can be executed with environmentally-friendly vehicles such as cargo bikes or electric vans. In addition to the huge investment costs for tunnel boring, this last-mile delivery option faces a challenging operational problem, which is the synchronization of goods arrivals at a capacity-restricted inner-city hub with the delivery tours. We formulate a basic two-echelon optimization problem that captures all main operational challenges of this synchronization task, derive suitable optimization procedures, and apply them in a computational study. The latter explores the interdependence between tunnel throughput, storage capacity at the hub, and vehicle capacity of the delivery person and their impact on delivery performance. An additional benchmark test with conventional truck-based deliveries shows that the reductions of carbon dioxide emissions promised by a cargo tunnel come for the price of excessive cargo bike traffic, especially if large urban regions are to be serviced from a micro hub.