Enabling Moving Target Hand-off in GPS-Denied Environments

ABSTRACT: Future conflicts in contested environments will require coordination between teams of manned and unmanned platforms performing ISR tasks. Coordinating ISR tasks is much more challenging in these contested conditions as GPS cannot be reliably depended upon. Consider the ISR task of handing off a tracked target from one platform to another. Reliably performing this task requires at a minimum an estimate of the relative pose (change in 3D position and orientation) between the two ISR sensors. In the GPS-enabled case, this is trivial to compute from the current GPS-INS position/attitude estimates on each of the platforms. In the GPS-denied case, however, this task becomes much harder.. Under this STTR we propose a joint effort between Systems & Technology Research and academic partner West Virginia University to develop a solution to GPS-denied relative pose estimation. Our proposed approach exploits data from multiple sensors, including imagery and inertial measurements, in a joint fusion framework for modeling and inferring relative pose in the absence of GPS. We will evaluate our approach in the GPS-denied, cross-modal target handoff application, and compare our performance with the traditional GPS-enabled case. BENEFIT: The technology developed on this program provides a statistically sound foundation for establishing a common coordinated frame between two platforms, thus enabling reasoning about coordinated tasks (with uncertainty) without GPS. This has wide applicability to any coordinated military activity in contested environments, including ISR tasks like target handoff as well as joint/cooperative navigation. Similarly, it has non-military application to any coordinated robotics application where GPS is unreliable, such as search-and-rescue.