Project Summary

We propose the demonstration of a simple detect and avoid algorithm onboard a drone using augmented reality (AR). The aim is to conduct constrained sense and avoid tests where a hovering drone will take evasive action to a virtual aircraft. The threat will be classified by type of object and collision urgency based on computer vision and deep learning techniques. The individual components arerelatively mature, and this project proposes a system of systems approach. Firstly, it will take place in a simulation environment with hardware in the loop (HITL) for Detect And Avoid (DAA) processing and flight control. These will be extended to a drone in flight using a ground station to stage the AR simulation.