There is a need for robust guidance, navigation, and control (GNC) capabilities for individual and cooperative teams of unmanned air vehicles (UAVs) in a wide array of environments. In the absence of consistent GPS feedback, inertial navigation systems (INS) inevitably suffer from degraded performance. For this reason, AFRL is interested in working with researchers on a wide array of GPS-denied INS approaches, including leveraging geo-referenced databases, non-geo-referenced approaches, and particularly cooperative approaches (possibly leveraging inter-agent measurements such as range, etc.).  

 

Proposed research applied to single agent INS approaches should improve system robustness over current state of the art, either through improvements to existing techniques or through alternative and complementary aiding strategies. Additionally, improving estimation quality through integrated guidance and/or control is of interest. It has been shown that trajectory choice has a dramatic effect on some GPS-denied INS approaches (e.g. visual SLAM). A formal approach is desired to identify effective guidance/control laws that address both the overarching navigation goal, but also minimizing or appropriately limiting INS uncertainty in the process. 

 

Proposed research in multi-agent, networked navigation should expand the current estimation paradigm and may involve significant departure from current distributed estimation literature. Proposed approaches must consider communication bandwidth limitations, likely avoiding centralized solutions. Ideally, cooperative INS approaches will gracefully degrade/recover in the presence of communication dropouts, delays, and updates in network topology. Researchers are also encouraged to consider the use of inter-agent measurements such as range to aid the INS and/or guidance laws. Again, guidance/control laws which benefit the networked agent’s INS accuracy is desired; in this case, navigation accuracy and network topology become paramount as both the bandwidth for information exchange and the quality of information will affect INS performance.  

 

Researchers will have access to AFRL/RW’s Autonomous Vehicles Lab (AVL) located at the University of Florida’s (UF) Research Education and Engineering Facility (REEF), in Shalimar, Florida. The AVL is intended to provide proof-of-concept hardware capabilities (small ground and multi-rotor vehicles) and/or simulation capabilities for developmental GNC efforts. The lab consists of a 25’x40’x12’ flight volume, motion capture (for development and truth reference), vehicles, several indoor sensors, onboard processing, base-stations, and simulation workstations.  

 

The AVL is staffed with a full-time research engineer capable of assisting with the use or modification of existing lab capabilities to expedite proof-of-concept efforts. The majority of our existing lab capabilities are implemented in the Robot Operating System (ROS) with associated algorithms coded using either C++ or Python. The AVL enjoys a wide range of permanent and visiting researchers each year, including postdocs, visiting faculty, summer graduate interns, undergraduate interns, and visiting AFRL researchers, creating a multi-disciplinary and interactive GNC development environment.

 

key words

Estimation; Navigation; Guidance; Control; Cooperative Navigation; Multi-agent Navigation; GPS-denied; Robust Navigation; Cooperative Estimation

Citizenship:  Open to U.S. citizens

Level:  Open to Postdoctoral and Senior applicants