Artificial Intelligence Enhanced Perimeter Security for Detecting Wildlife Intrusion and Improving Uptime

Ellsworth Air Force Base (AFB) resides in the Black Hills of South Dakota, an area known for its  natural beauty and extraordinary wildlife. While such wildlife is valuable from an ecological perspective outside of the base, within Ellsworth AFB, wildlife can present hazards to equipment, personnel, and airway safety. The Federal Aviation Administration (FAA) estimates annual damage to United States (US) aircraft of approximately $400 million dollars due to bird strikes alone. Current methods of perimeter security require an Airman to manually monitor camera feeds. Monitoring can be tedious and can be limited to single or limited number of camera feeds at a time, with detection efforts requiring full and consistent attention of an Airman. Even adding only a second camera feed reduces available attention 50%. Artificial Intelligence (AI) continuous monitoring will be a force multiplier for the Airman monitoring video feeds in four primary ways. First, AI algorithms tire, and always apply the same level of analysis. Second, AI can monitor and combine multi-modal data (such as thermal, infrared, night vision cameras, or others) to improve detection robustness.  Third, the output of AI algorithms can be catalogued for future reference and analysis to assist in detecting systematic intrusion issues. Fourth, AI algorithms can classify and locate targets in a monitored video feed for rapid triage, bringing dangerous issues rapidly to an airman’s attention. Analytical AI (AAI) is a small business that specializes in AI, developing novel AI approaches designed to integrate with customer preferred software, and run rapidly on hardware appropriate to the application.  AAIs proprietary algorithms can be readily tailored to address this perimeter security need.  Technology development under this effort contributes to mission need fulfillment in the area of sensors and AI. AAI will use Ellsworth AFB as a use case for technology that will be useful not only for other military installations and AFBs (such as Tyndall AFB), but also commercially in the instance of identifying wildlife on commercial airfields, on game cameras, and perimeter security for identifying persons intruding into secure areas, persons carrying visible weapons, unattended baggage, and many other applications. The specific mission impact of this project on the Department of the Air Force (DAF) and Department of Defense (DoD) will include safer AFBs and reduced base downtime. The solution focuses on delivering correct information quickly to allow airman to mount an appropriate response.