A network design model for food ordering and delivery services

The remarkable developments in e-commerce offer more flexibility and alternative shopping experience to busy people, to such an extent that the best choice in the future for having a meal may be to have it delivered at home. Food ordering and delivery is not a recent business, but it scaled up in the last few years, thank to several successful "logistics apps" and online platforms that enabled online meal purchasing from a wide range of restaurants, promising fast and reliable deliveries of food within city limits. Those revolutionary, original and efficient takeaway services, such as Just-eat, Deliveroo, Deliver Hero and others, gained popularity both for their business success and for class actions related to the employment conditions. In this paper, we first analyse how do these emerging companies work, providing a special focus on the logistics aspects. In particular, we analyse the operations related to the order pick-up and delivery process, performed relying on large networks of self-employed contractors (i.e. bike drivers). Analysing the network, we recognise as a critical factor the location from where the carriers depart for performing the service. In fact, the carrier has a limited time to pick up and deliver the orders; thus, the distances between the carrier's departing point, the restaurants and the customers play a substantial role in determining the service offer. We formulate an optimization model that provides insights into the network design process. The model is an extension of the maximal covering location problem aimed to determine the location and the number of carriers' departing points. The objective function is to maximise the demand covered. This model can be further extended to include other service aspects.