Self-driving systems used in electric vehicles now only operate on paved road areas with clear road dividers and good physical barriers. When asked to move on unpaved roads, these systems continue to have difficulty operating. This causes the existing self-driving system not to be used for military purposes.
DARPA has announced the Robotic Autonomy in Complex Environments with Resiliency (RACER) project to develop a self-driving system that can operate on unpaved roads with performance as good as a human driver.
RACER was developed by DARPA in collaboration with Carnegie Mellon University, NASA's Jet Propulsion Laboratory, and the University of Washington. The first phase of testing was carried out earlier this year to test the obstacle avoidance system, and find the best route in the desert.
The next test is to see RACER's ability to climb steep, slippery areas and more complex obstacles. The algorithm's ability to solve real-time driving issues will be fully tested.
RACER is a continuation of the AMAS (Autonomous Mobility Appliqué System) project developed since 2014 which allows a group of trucks to move in a convoy in a self-driving manner to reduce risks when operating on the battlefield.