Flies, Data, Robots: Insect-Inspired Flight Strategies for Robust UAV Navigation in GPS Denied Environments
Loading...
Authors
Lopez, Austin Perry
Issue Date
2025
Type
Dissertation
Language
en_US
Keywords
Bio Inspired , Controls , Drone , GPS Denied , Observability , UAV
Alternative Title
Abstract
Unmanned aerial vehicles (UAVs) are increasingly called upon to inspect pipelines, track agricultural volatiles, and map disaster zones where Global Positioning System (GPS) signals are unreliable or actively jammed. Inspired by the active sensing strategies of flying insects, this dissertation explores how carefully choreographed flight trajectories can reconstruct lost odometry. I focus on a midsize UAV; however, the results extend to micro UAVs whose size or damage precludes stereo or depth cameras and limits the payload to an inertial measurement unit (IMU), a monocular camera, and a wind probe. Using nonlinear observability analysis, I first show how insects could estimate ground speed and wind speed with their antennae and deliberate maneuvers. Adapting these insights to my UAV, the “BIG BUG,” I designed insect-inspired trajectories such as cross-wind casting (sinusoidal paths), flew them in a motion-capture arena, and fused optic-flow, IMU, and wind sensor data in an estimator. The results demonstrate that these bio-inspired maneuvers improve velocity and wind-state estimates compared with naïve straight-line flight.