The project develops a fast optical guidance and navigation suite for small unmanned air systems (SUASs). The suite integrates novel sensors, processing and behaviors inspired from biology together with state-of-the-art engineered sensors, such as lightweight LIDAR and broad-spectrum vision sensors. Specification development are established to meet and surpass benchmarked performance targets by the end of Phase II of 98% detection probability of potentially interfering airborne objects, while keeping a target weight under eight pounds and assuring obstacle avoidance capabilities to ensure safe flight to complete mission objectives. The low cost, low SWaP optical guidance and navigation system take origin in bio-inspired navigation behaviors refined with representations and control algorithms adapted to available sensors and processing platforms using machine learning and deep learning techniques. The system provides, at low budget, a complete autonomous navigation system with take-off, landing, mapping and collision avoidance capabilities to enable a small, low-weight SUAS affordable, robust and unassisted airborne navigation in various flight environments. Finally, the navigation suite architecture is designed to be open and flexible for rapid changes in sensors and algorithms.
