Physical Optics Corporation SBIR Phase I Award, September 2019

To address the Navy’s need for artificial neural networks (ANN)-centric reinforcement learning (RL) algorithms that provide unmanned aerial vehicles (UAVs) with the capability to autonomously conduct flight from takeoff to landing, Physical Optics Corporation (POC) proposes to develop a new Reinforcement Learning Algorithms for Unmanned Aerial Vehicles (REALUAV) algorithm suite. It is based on a unique integration of state-of-the-art deep RL algorithms and techniques that enable the REALUAV agent to learn and master a set of diverse tasks sequentially or simultaneously and, given a flight profile, accomplishes a mission within an unknown environment optimally. REALUAV offers autonomous operation of UAVs, modifiable in real-time by a human-in-the-loop, which directly address the PMA268 Navy Unmanned Combat Air System Demonstration program requirements. In Phase I, POC will design, develop, and demonstrate the feasibility of REALUAV by reaching technology readiness level (TRL)-3. In Phase II, POC plans to develop a REALUAV prototype algorithm suite reaching TRL-5/-6.