Mobile Wildlife Surveillance System

The Air Force Base has a requirement, based on the Endangered Species Act, the Sikes Act, and the Migratory Bird Treaty Act, to maintain the natural environment near their facilities, while still meeting military mission objectives. We propose the creation and deployment of a remotely controlled or fully automated robotic (operator choice) mobile wildlife surveillance system (MWSS) or in support of managing natural resources. The system will be comprised of a small rugged robot that drives a search pattern based on a map and detected obstacles, and is able to collect video (visible light, IR, and perhaps thermal), audio, and weather related (e.g. temperature, pressure, and airspeed) data using a suite of relevant sensors. The sensor data will be recorded and analyzed using deep learning software trained to recognize the relevant fauna and animal dwellings. The recorded data will be tagged with the identification information, and if the confidence is high enough, the robot will remotely notify experts, allowing teleoperation (if desired) of the robot and the sensors in order to verify or refute the automated detection. This verification/refutation can then subsequently be used to improve the identification process via more training