CHARLES RIVER ANALYTICS, INC. — Department of Defense SBIR Phase I: A214-041

Description

While significant progress has been made toward developing robust obstacle detection and avoidance technologies for Unmanned Air Vehicles (UAVs) and Unmanned Ground Vehicles (UGVs), tethered systems face unique challenges that are not addressed by existing solutions. Charles River Analytics proposes to design an architecture and solution for Passive Ranging, Obstacle Tracking, and Efficient Coordination of Tethered Systems (PROTECTS) that enables safe, on-the-move operation of a Te-UAS across diverse environments and usage contexts. The proposed system integrates commercial off-the-shelf (COTS) sensors and computing hardware, and builds on recent advances in deep learning-based monocular depth estimation along with specialized obstacle detection methods to provide a fully passive, high-resolution sensing solution supporting safe Te-UAS operations. To ensure that future Te-UAS systems will be able to take advantage of the rapid improvement of relevant sensors and algorithms, the PROTECTS architecture will be designed following a Modular Open Systems Approach (MOSA), enabling plug-in replacement of key modules by future third-party developers.