CAREER: Radar-based Perception and Navigation in Visually Degraded Environments

University Of Colorado At Boulder, Boulder CO

Investigators

Abstract

Autonomous field robots perform increasingly complex tasks under uncertain and dynamic environmental conditions. These robots often rely on sensors like cameras and lidar for perception, but these sensors are hindered in environments where vision is degraded, such as smoke-filled rooms or dust-blown construction sites. This Faculty Early Career Development (CAREER) project supports research that explores the use of millimeter-wave radar, a technology capable of operating effectively in poor visibility, to enhance robot perception and navigation. Millimeter-wave radar can penetrate through obstructions in vision and function irrespective of lighting conditions, making it ideal for tasks in search and rescue, firefighting, and construction. This research has the potential to significantly advance robotic autonomy in challenging environments, benefiting society by improving operational efficiency in critical industries and contributing to the education and diversification of the future workforce in robotics. This research project aims to develop algorithms for interpreting and processing radar data, translating raw signals into a comprehensive understanding of the environment, including object detection, classification, localization, and mapping through feature learning on dense radar data. The integration of radar-based perception capabilities with robotic navigation will create a robust system for navigating through VDEs. Research activities will include developing high-precision metric localization and mapping techniques, learning representations for direct navigation through learning over partially observable environments, and validating these techniques on experimental testbeds. By leveraging advancements in millimeter-wave radar sensors, this project will enhance robotic autonomy and contribute to the broader scientific community by providing new tools and data for further research. The outcomes will be integrated into educational programs to train the next generation of scientists and engineers, fostering innovation in the field of robotics. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.