How can we build trustworthy unmanned vehicle systems?
How can we enable reliable operation in adverse weather such as snow, rain, and fog?
How can we handle challenging environments including rough terrain, maritime domains, and underwater settings?
How can we quantify system reliability?
These questions drive our research on uncertainty-aware perception for field robots.
We develop perception methods that explicitly model uncertainty, enabling robotic systems to operate more robustly, safely, and reliably in complex real-world environments.