Frontier Exploration for Disaster Management UAVs

Disaster management UAVs are designed to aid mainly in the rescue by providing important information to the first responders from a distance while avoiding difficulties of terrain. Autonomous UAVs allow for elimination of human error in navigation. While most indoor exploration strategies aim to minimize total exploration time, that is not suitable for disaster scenarios where a communication channel between the UAV and the first responders is present. In such cases, the goal is to deliver critical information such as map of the environment and human locations to the first responders amid the journey. In this paper we present a frontier exploration strategy adapted for disaster management UAVs in indoor scenarios. The proposed algorithm treats detected humans as clusters that are prioritized for exploration. The UAV is capable of completing the exploration in reasonable time while prioritizing the discovery of humans.