Refine
Year of publication
- 2020 (1)
Document Type
- Article (1) (show_all)
Language
- English (1)
Has Fulltext
- yes (1)
Is part of the Bibliography
- no (1)
Keywords
- Dynamic programming (1)
- Flexible product (1)
- Revenue management (1)
- e-commerce (1)
Institute
- Mercator Endowed Chair of Demand Management & Sustainable Transport (1) (show_all)
In e-commerce, customers are usually offered a menu of home delivery time windows of which they need to select exactly one, even though at least some customers may be more exible. To exploit the exibility of such customers, we propose to introduce exible delivery time slots, defined as any combination of such regular time windows (not necessarily adjacent). In selecting a exible time slot (out of a set of windows that form the exible product), the customer agrees to be informed only shortly prior to the dispatching of the delivery vehicle in which regular time window the goods will arrive. In return for providing this exibility, the company may offer the customer a reduced delivery charge and/or highlight the environmental benefits. Our framework also can accommodate customized exible slots where customers can self-select a set of regular slots in which a delivery may take place.
The vehicle routing problem (VRP) in the presence of exible time slots bookings corresponds to a VRP with multiple time windows. We build on literature on demand management and vehicle routing for attended home delivery, as well as on exible products. These two concepts have not yet been combined, and indeed the results from the exible products literature do not carry over directly because future expected vehicle routing implications need to be taken into account. The main methodological contribution is the development of a tractable linear programming formulation that links demand management decisions and routing cost implications, whilst accounting for customer choice behavior. The output of this linear program provides information on the (approximate) opportunity cost associated with specific orders and informs a tractable dynamic pricing policy for regular and exible slots. Numerical experiments, based on realistically-sized scenarios, indicate that expected profit may increase significantly depending on demand intensity when adding exible slots rather than using only regular slots.