Utilizing ML Algorithms to Track and Identify UAS Threats

Special Operations Command is responsible for many of our nation's most critical, no-fail missions, yet the rapid rise of Unmanned Aerial Systems (UAS) is forcing rapid adaptation in the ways these forces can detect, track, and characterize these threats. Our vision is to explore the art of the possible, pairing portable commercial LiDAR sensors with Computer Vision and Deep Learning algorithms to facilitate the identification, localization and pursuant tracking of UAS via a portable, dismounted system. Our feasibility design provides a robust framework to evaluate the best machine learning solutions for object detection across a variety of UAS types, the ability to maintain custody of UAS tracks, and a system design comparative analysis for LiDAR and ML solutions that provide range maximization, detection accuracy, integration with radar, and integration with active mitigation measures. Our solution is backed by experts who bring real-world combat experience, software and machine learning expertise from Silicon Valley, and peer-reviewed research specialization. We have extensive experience designing and developing complex deep learning solutions both commercially and for defense.