UAV Based Survivor Search during Floods

This paper addresses the problem of searching for survivors in flood affected regions. These search missions are carried out by UAVs (Unmanned Aerial Vehicles) in order to search faster and cover larger distances. The proposed search algorithm receives information regarding survivor location from the ground based observers. The observers report the location of survivors when the survivor is within its field of view. In such a situation when the location information about the survivor is received by the UAV, the survivor might have moved to a different location when the UAV tries to go to the last spotted location to search. Hence, it is necessary to have a prediction algorithm to estimate the future position of the survivor. The algorithms implemented in this paper is a weightage based approach and a truncated weight based approach to predict the future location of the survivor. The performance of these algorithms are compared with a tradition lawnmower search pattern. These algorithms are implemented by taking a real life flood scenario and the effectiveness and efficiency of the implemented algorithm is shown by an extensive Monte Carlo simulation.