Creating ground vehicles that guide themselves is the most challenging robotic work. SARA has developed a
specific capability for the design and fabrication of autonomous vehicles and in the integration of sensor
systems onto autonomous ground and air vehicles.
SARA technology in real-time video image processing allows use of monocular and stereo video cameras for
vehicle navigation, control, and obstacle avoidance. Such systems can operate with totally passive sensors
in a light-weight, low-power configuration.
SARA technology in controls and sensor systems comes from our experience with remotely controlled robots, and
from development of vehicles for the DARPA Grand Challenge, an off-road autonomous vehicle race across the desert.
SARA has also developed navigation systems that operate in GPS-denied environments using multiple absolute and
relative sensor systems, combined with GIS data, integrated under an Extended Kalman filter.
UAV Ground Operations System (UGOPS)
This program for the Air Force Research Labs, tackles the problem of operated-assisted autonomous movement of
unmanned aerial vehicles on the ground. How do you move about on taxi-ways and ramps, avoiding obstacles and
waiting in line behind other manned and unmanned aircraft. How do you do this on a poorly mapped airfield
without GPS? Our solution uses video cameras on the aircraft to determine the edges of taxiways, and to look
out for obstacles in the path. Inertial sensors, updated with video-sensed edge data, odometer data, and digital
compass data provide sufficient sensory input to safely conduct operations.
SARA's UAV Ground Operations System
- Robust navigation using redundant sensors and GPS
- Vision-based processing provides the greatest sensor data and bearing accuracy in a lightweight, low-cost sensor
- Integrated navigation, control, and sensor processing