Rescue Boat Path Planning in Flooded Urban Environments

EasyChair Preprint 1316

Abstract

Mapping and path planning in disaster scenarios is an area that has benefited from aerial imaging and unmanned aerial (UAV) and surface vehicles (USV). Nowadays, there are many application areas of UAV and USV which consist of environmental monitoring, victim recovery and river mapping in the purpose of rescue operations. There are many challenges for flood rescue operations with existing systems where the time response is critical. The environment is completely changing under flood conditions and it causes a dangerous environment for victims and rescue operators. Rescue boats are widely used for searching and rescuing the victims in the flooded areas. However, rescue boats have a limited view while searching and rescuing the victims. By using UAVs, an aerial image can be taken off the flooded environment and this aerial image can provide global information such as the location of victims and landmarks. The rescue operations can be organized based on a generated flood environment map by using this global information. Here we propose a ground map generation and path planning algorithms, which makes use of aerial imaging provided by a UAV in flooded urban environments. Based on the generated ground map, we used the concept of mobile robot path planning to represent our proposed approach in rescue boat path planning. In this purpose, A*, GA and PRM path planning algorithms are used and analyzed to find near-optimal paths for rescue boat between initial and target locations. Experiments are performed to evaluate the performance of the system algorithms to find out the most suitable algorithm in flooded urban environments.

Keyphrases: Flood rescue operations, Rescue boat, Unmanned Aerial Vehicle, map building, path planning

@booklet{EasyChair:1316,
  author    = {Mehmet Ozkan and Luis Rodolfo Garcia Carrillo and Scott A. King},
  title     = {Rescue Boat Path Planning in Flooded Urban Environments},
  howpublished = {EasyChair Preprint 1316},
  year      = {EasyChair, 2019}}