Who moves the locker? A benchmark study of alternative mobile parcel locker concepts.

Stationary parcel lockers have established as useful delivery options, especially for those households where nobody is at home during typical parcel delivery times. However, having to move toward a locker, maybe after a tiring workday, can be inconvenient for customers. Mobile parcel lockers, especially when driving autonomously, can be positioned closer to the customers and thus seem as a logical next evolutionary step to reduce customer inconvenience. Until autonomous driving is finally realized, however, mobile parcel lockers depend on human-driven vehicles to reposition them, and this paper compares optional locker–vehicle–driver setups. In the most restrictive case, a parcel locker is fixedly mounted onto a vehicle and equipped with a dedicated driver. But a human driver can also be in charge of multiple lockers, so that the driver must travel, e.g., via public transport, between different lockers in order to reposition them. There are also first concepts without a fixed coupling of lockers that are only loaded onto their vehicles. Hence, a vehicle equipped with an automated handling mechanism can subsequently reposition multiple mobile locker modules between different parking positions. Based on the assumption that a given set of customers is to be serviced with a predefined service level, this paper provides a flexible heuristic multi-stage optimization approach that minimizes the total costs associated with each of these concepts. This algorithm is applied in a benchmark study to compare the alternative mobile parcel locker concepts, and our results reveal substantial differences among them. For instance, we show that mobile locker modules are very effective to provide better customer service (i.e., smaller walking distances and longer parking durations) at low cost • We benchmark five different mobile locker concepts against stationary lockers. • Mobile lockers with autonomous driving and human drivers are compared. • A heuristic framework that can be applied to all locker concepts is introduced. • Service expectations and their impact on locker resources are investigated. [ABSTRACT FROM AUTHOR]