The US Defense Advanced Research Project Agency (DARPA) has begun a second field experiment to assess unmanned combat vehicles in realistic driving situations at Camp Roberts, California.
The experiment is being conducted under DARPA’s Robotic Autonomy in Complex Environments with Resiliency (RACER), in which driverless vehicles equipped with off-road autonomy technologies are being developed to match the skills and speed of human drivers.
“We are after driverless ground vehicles that can maneuver on unstructured off-road terrain at speeds that are only limited by considerations of sensor performance, mechanical constraints, and safety,” DARPA RACER Program Manager Stuart Young said.
“At a minimum, the program goal is software performance that allows off-road speeds on par with a human driver.”
DARPA works on RACER with experts from NASA’s Jet Propulsion Laboratory, Carnegie Mellon University, and the University of Washington.
RACER Experiments
Experiment 1 was held in Fort Irwin, California, in March and April 2022. It included six courses of combat-relevant terrain where the robot vehicles carried out 40 autonomous drives for around 2 miles (3.2 kilometers) at under 20 miles (32 kilometers) per hour.
During the experiment, the testers faced challenges in the vehicles’ ability to identify, classify, and avoid obstacles when exceeding the preferred speed.
Fort Irwin’s near-desert environment provided obstacles and hazards that could have damaged the vehicles.
In Experiment 2, the robot vehicles’ perception algorithms were tested, which help them to navigate challenging terrain while maintaining control. The site was engineered with features including larger and steeper hills.
“Since the first experiment, teams have been working to improve perception of the environment and planning navigable routes through development of new autonomy algorithm technologies,” Young explained.
“The DARPA-provided RACER fleet vehicles being used in the program are high performance all-terrain vehicles outfitted with world-class sensing and computational abilities, but the teams’ focus is on computational solutions as that platform encounters increasingly complex off-road terrain.”