The Stochastic Vehicle Routing Problem for Large-scale Emergencies

In this paper, we are interested in routing vehicles to service large-scale emergencies, such as natural disasters and terrorist attacks. We describe the specifics of routing vehicles in a large-scale emergency and decompose the problem into two stages: a planning stage and an operational stage. In the planning stage, we aim to generate the routes well in advance of any emergencies; in the operational stage, we decide the delivery quantity with the planned routes and make adjustments to the routes if necessary at the time of the emergency when more information is revealed. We propose and discuss mathematical formulations and solution approaches for both stages. Lastly, we demonstrate the effectiveness of our formulations in developing robust routes through numerical experiments.