Demand and Capacity Management in a Stochastic Dynamic Pickup and Delivery Problem with Crowdsourced Resources

In this study we examine the situation of a consortium of brick-and-mortar businesses operating an e-portal to collect customer orders and provide same-day delivery as a dynamic and stochastic pickup-and-delivery problem in which deliveries are performed with a mixed fleet of vehicles composed of dedicated vehicles and crowdsourced resources. Forms of delivery capacity vary with respect to the level of information and control that the dispatcher has on the availability and behavior of the courier. Taking into account the specific nature of the different types of delivery capacity, we present an approximate dynamic programming approach to manage the same-day delivery. We present computational results for different scenarios of demand.