UAV Navigation

UAV navigation studies how drones estimate their position, understand nearby space, and choose motions to reach goals while avoiding obstacles and respecting vehicle limits. State estimation commonly uses visual-inertial or LiDAR-inertial odometry, sometimes combined with satellite imagery, semantic maps, or other sensors when GNSS is unavailable. Planners range from model predictive control, graph search, and trajectory optimization to reinforcement learning and vision-language-action models that follow natural-language goals; safety filters and control barrier functions are often used to constrain motion. A central challenge is making these systems reliable in unfamiliar, dynamic conditions while balancing safety, speed, energy use, onboard computation, and transfer from simulation to real flight.