Designing logistics systems for home delivery in densely populated urban areas

•We show urban delivery systems in practice are cost-effective under high density.•We introduce a simple aggregation concept for demand growth and variation.•We provide insight by a worst-case analysis for specific topologies. To deliver to customers in densely populated urban areas, companies often...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Transportation research. Part B: methodological 2018-09, Vol.115, p.95-125
Hauptverfasser: Huang, Yixiao, Savelsbergh, Martin, Zhao, Lei
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:•We show urban delivery systems in practice are cost-effective under high density.•We introduce a simple aggregation concept for demand growth and variation.•We provide insight by a worst-case analysis for specific topologies. To deliver to customers in densely populated urban areas, companies often employ a two-echelon logistics system. In a two-echelon logistics system, the starting point for goods to be delivered in the urban area is a fulfillment center or city distribution center. From there, the goods are first transported to a satellite, from where they are delivered to their final destinations. To simplify operations, companies, in practice, often restrict the route choices at one or both of the echelons. We study the impact on the delivery cost of such restrictions and conclude that when the number of orders to be delivered is large and the location density of delivery addresses is high, the impact is likely to be small, and the operational benefits probably outweigh the cost increases. To more easily accommodate delivery volume growth and to more effectively handle day-to-day delivery volume variations, we introduce a simple aggregation concept, which leads to quality improvements without affecting operational simplicity. We provide further insight by means of a worst-case analysis for specific geographic topologies.
ISSN:0191-2615
1879-2367
DOI:10.1016/j.trb.2018.07.006