Automated parcel lockers location problem: a numerical experiment for Turin’s urban area

The significant and constant spike in e-commerce purchases experienced in recent years has generated more pressure on Logistics Service Providers’ urban activities, with negative consequences on costs and negative externalities (i.e. environmental impacts and traffic congestion). In this context, home deliveries are some of the major sources of inefficiencies in last-mile delivery systems, due to several reasons such as high order frequency, disperse demand, and nonnegligible probability of missed delivery. Consolidating final users’ demand in Automated Parcel Lockers (APL) represents an effective solution to the operational strains entrenched in last-mile deliveries. However, this solution requires significant investment by private operators and compels space organizers such as public authorities to grant the usage of portions of public space wherein APLs are installed. An accurate ex-ante appraisal of the variables involved is thus needed in order to discern the overall impact of APL on the urban setting. Hence, this study proposes a solution to the Parcel Lockers network design problem. In particular, the aim of the approach is to minimize the lost demand percentage and the number of APLs installed. The dimensions of the APLs are also assessed by the proposed algorithm in a second step. Input variables and parameters of the model are identified through an online survey submitted to final users located in the city of Turin, Italy. Case-scenarios are simulated via 1000 random demand points according to the real population distribution, potentially covered by 33 APL locations located inside the 90sqkm urban area. The purpose of this work is to draw implications for last-mile private actors, urban space planners and policy makers