All-weather perception for man-portable robots using ultra-wideband radar

Autonomous man-portable robots have the potential to provide a wide range of new capabilities for both military and civilian applications. Previous research in autonomy for small robots has focused on vision, LIDAR, and sonar sensors. While vision and LIDAR work well in clear weather, they are seriously impaired by rain, snow, fog, and smoke. Sonar can penetrate adverse weather, but has limited range outdoors, and suffers from specular reflections indoors. For the Daredevil Project, we have investigated the use of ultra-wideband (UWB) radar to provide obstacle detection capabilities for man-portable robots. Our research shows that UWB radar can effectively penetrate adverse weather, including dense fog, and detect obstacles that would be undetectable by vision or LIDAR under the same conditions. We have developed filtering algorithms that process the raw radar returns to eliminate reflections from ground clutter and make obstacles easier to detect. We have tested this system on an iRobot PackBot equipped with both UWB radar and LIDAR, and we have demonstrated how UWB radar can be used for obstacle detection in obscured environments.