An Unmanned Aerial System for Prescribed Fires

Prescribed fires can lessen wildfire severity and control invasive species, but some terrains may be difficult, dangerous, or costly to burn with existing tools. This thesis presents the design of an unmanned aerial system that can ignite prescribed fires from the air, with less cost and risk than with aerial ignition from a manned aircraft. The prototype was evaluated in-lab and successfully used to ignite interior areas of two prescribed fires. Additionally, we introduce an approach that integrates a lightweight fire simulation to autonomously plan safe flight trajectories and suggest effective fire lines. Both components are unique in that they are amenable to input from the system’s sensors and the fire crew. A preliminary study confirms that such inputs improve the accuracy of the fire simulation to better counter the unpredictability of the target environment. Advisors: Carrick Detweiler and Sebastian Elbau