Target handoff in GPS-denied environments
The Handoff Problem: When multiple unmanned aerial systems (UAS) coordinate their efforts to perform surveillance and target tracking missions, they need to ensure that they are tracking the same target of interest. In many cases, the tracking responsibility may need to be handed off from one UAS to another.
Existing solution: If GPS signals are available, such a handoff procedure can be achieved by the tracking UAS first geo-locating the target and then sending the global coordinate of the target to the handoff UAS. The handoff UAS can use the received global coordinate of the target to point its sensor at the target.
Challenges in GPS-denied environments: However, GPS signals are subject to intentional jamming or unintentional interference and blockage in cluttered urban environments. In GPS-denied or degraded environments, such a handoff procedure faces several challenges, such as global information of the target being not available.
Our approach: We propose an approach to the target handoff problem in GPS-denied environments and address the observability questions associated with this approach. In this approach, we assume that each UAS has an Inertial Navigation System (INS) that can be used to estimate its roll and pitch angles without GPS aiding. By allowing the tracking UAS to transmit the estimated roll and pitch angles to the handoff UAS, we reduce the 6 degree-of-freedom (DOF) relative pose estimation to relative position and relative heading (relative yaw) estimation.
We relax the common assumption that vehicles exchange their linear and angular speed information for relative localization and consider the case where the airspeed and turn rate of the tracking UAS are not available to the handoff UAS. In this case, we establish observability properties of the relative position and the relative heading given different types of sensor measurements. These properties demonstrate the minimum data requirements for our target handoff approach and reveal redundancies inherent in relative localization. Nonlinear estimators using a bank of EKFs are designed to estimate relevant information.
Two figures below illsutrate the handoff process: The handoff UAS estimates the relative position and heading between the two aircraft with only bearing angles, compute the target location, and points its camera to the target based on the computed target location.
Initially, the tracking UAS is tracking a ground target and the handoff UAS flies towards the tracking UAS with no state information of the target or the tracking UAS.
At the end, the handoff UAS captures the target in the field of view after the relative state estimates converge.
Publications:
[J10] H. Bai and Randy Beard. Relative heading estimation and its application in target handoff in GPS-denied environments. Submitted to the IEEE Transaction on Control System Technology, 2017.
[C25] A. Awan and H. Bai. Relative heading estimation under constant disturbances. In the Proceedings of the 2017 American Control Conference, Seattle, WA, 2017.
[C22] H. Bai and Randy Beard. Relative heading estimation for target handoff in GPS-denied environments. In the Proceedings of the American Control Conference, 2016.