Unmanned Aerial Vehicle Mapping with Semantic and Traversability Metrics for Forest Fire Mitigation
Abstract
Climate change and other factors are causing increasingly frequent and intense forest fires worldwide. Robotics systems can improve the feasibility of prevention and mitigation efforts. In this work, we propose an unmanned aerial vehicle that can map a forest region to allow an unmanned ground vehicle to autonomously clear fuel to prevent the spread of fire. We developed a multi-sensor payload consisting of cameras, LiDAR, and GPS with onboard processing. We also implement a SLAM system to understand the 3D structure of the environment, a semantics system to identify fuel and other features in the environment, and a traversability system that predicts which region a UGV can navigate. This approach provides a 3D map of the environment and geo-registered maps describing the locations of fuel and traversable regions. We validate our method with preliminary field trials and show that this is a promising approach.
BibTeX
@workshop{Russell-2022-137685,author = {David Russell and Tito Arevalo and Chinmay Garg and Winnie Kuang and Francisco Yandun and David Wettergreen and George Kantor},
title = {Unmanned Aerial Vehicle Mapping with Semantic and Traversability Metrics for Forest Fire Mitigation},
booktitle = {Proceedings of ICRA 2022 Innovation in Forestry Robotics: Research and Industry Adoption},
year = {2022},
month = {May},
}