Field-based Autonomous LiDAR Control for Obstacle Navigation (FALCON)

Unmanned Aerial Vehicles (UAVs) are increasingly used for surveillance and monitoring in urban environments, where effective obstacle avoidance is critical for safe and efficient operations. This paper presents a LiDAR-based obstacle avoidance system using the 3D Vector Field Histogram Plus (3D VFH+) method, implemented on the Field-based Autonomous LiDAR Control for Obstacle Navigation (FALCON), an in-house UAV developed at Ulster University. The study focuses on UAV navigation around the university campus, specifically York Street, where buildings pose significant challenges for real-time obstacle avoidance. A digital twin model of FALCON is used for simulation and validation, integrating 3D VFH + with a proportional integral derivative (PID) control strategy optimised using particle swarm optimisation (PSO), genetic algorithm (GA) and Mayfly algorithm (MA). The system is tested in a simulated environment including the Ulster University Belfast (UUB) campus, comparing different PID tuning methods in terms of path deviation, stability, and obstacle clearance. Results demonstrate that PSO and GA-optimised PID controllers provide smoother trajectories and improved stability, ensuring safe UAV navigation in complex urban settings. This research contributes to the development of robust UAV obstacle avoidance strategies for urban surveillance and monitoring applications.