Key points are not available for this paper at this time.
Solving Vehicle Routing Problems (VRPs) is one very important step for modeling and simulating urban freight transportation and especially urban last-mile delivery. It is observed in several studies, that tour plans consist of a very high number of vehicles needed because vehicles only drive one tour. For the next tour, a new vehicle with new fixed costs is needed. This also starts in the beginning of its temporal availability. This article intends to answer the question: how tour plan and fleet size changes when vehicles can reload at the depot and run another tour within its temporal availability. The possible effects are shown in two studies for Berlin: (i) food retailing delivery and (ii) parcel delivery for one district of Berlin. The studies have different designs and policy cases to see the effects under different situations. The present research shows that reloading within the delivery tour leads to a significant reduction of vehicles needed and therefore to lower costs for the carrier. If one tour is not limited by the vehicle’s capacity but by the vehicle’s temporal availability no saving can be observed. This was the case for parcel delivery with light parcel delivery trucks. When using smaller vehicles, e.g. cargo bicycles for last mile delivery from local city hubs, vehicle capacity is the limiting factor and therefore also significant saving can be observed. As another consequence it is possible to get rid of artificially inserted breaks in time-windows for more realistic traffic. The article also critically addresses the problem of increasing computation time.
Martins-Turner et al. (2019) studied this question.