Unmanned Systems R&D
Developing advanced perception systems to determine terrain and roadway traversability and detect objects and obstacles is a core competency of Southwest Research Institute (SwRI). Using multi-modal sensor arrays (Cameras, LIDAR, RADAR, etc.) SwRI develops high-performance, low-cost perception systems to support various unmanned systems, driverless cars, and unmanned ground vehicles (UGVs).
SwRI has developed high-performance relative and absolute localization algorithms that allow unmanned systems to operate in GPS denied or degraded environments. These algorithms fuse relative motion sensors, vehicle and visual odometry, and absolute reference signals in a low-cost package that can yield high accuracy position correction.
Using modular advanced perception and localization algorithms to understand the environment, SwRI has implemented numerous vehicle control/path-planning algorithms. Distributed and real-time control systems operating reactively on-road and deliberatively off-road. SwRI also heavily utilizes the NIST 4D/RCS approach to behavior-based control.
Examples of SwRI Unmanned Systems (UGV) Programs
- Mobile Autonomous Robotics Technology Initiative (MARTI®) program
- Small Unit Mobility Enhancement Technology (SUMET)
- Intersection Navigation Algorithms for Automated Vehicles
- Negative Obstacles Detection
- Pedestrian Detection Technology
- Natural Language Control of UGVs
- Gesture Recognition for UGV Control
- Relative and Absolute Localization
- DBW System Analysis
- Next-Gen ITS Technology—Video
- Active Safety & Driverless Vehicles—Video
- MARTI|XE Data Collection in Thun, Switzerland–March 2014—Video
- Ranger: Localization System Using a Ground-Facing Camera—Video
- Waypoint Splining for Autonomous Vehicle Following
- Location and Motion Estimation Using Ground Imaging Sensor
- Optical State Estimation and Simulation Environment for Unmanned Aerial Vehicles