Vehicle Routing to Minimize the Average Delay for Collecting the Disaster Damage Information

When a large-scale disaster happens, the emergency disaster response headquarters (HQ) is set up to collect the dis-aster damaged information and conduct an appropriate disaster management plan. However, if the high-speed communications infrastructures are unavailable, such information should be brought to HQ by other means such as mobile vehicles. In this paper, therefore, we consider the scenario in which such information in a region are monitored and brought to HQ by patrolling vehicles equipped with cameras, mics, and other sensors. The patrolling vehicles can return to HQ multiple times on the way to drop the monitored information. A metric, Information Delay-time Product (IDP), is defined to represent the average delay time for information collection, i.e., the product of the fraction of the information and its delay time in delivering to HQ. By leveraging Eulerian circuits, we can systematically search good traveling routes for vehicles to incrementally collect the information along all streets. On a grid map as a town's street network with two vehicles, we experimentally show that an appropriate number of returns to HQ with balanced-sized intervals is necessary to minimize the IDP.