Intel to develop simulation for DARPA’s autonomous terrain vehicle program

Within the scope of the RACER program developed by the Advanced Defense Research Projects Agency (DARPA) of the US Department of Defense, Intel offers advanced simulation solutions for off-road autonomous land vehicles. will develop.

A two-stage study will be developed

With the system, called RACER-Sim, the focus will be on significantly reducing the development cost and closing the gap between simulation and the real world.

German Ros, executive director of Intel Labs: “We are proud to join RACER-Sim to continue contributing to the next frontier of off-road robots and autonomous vehicles. Versatile and open to accelerate progress in off-road robots for any environment and condition. We brought together a team of renowned experts from the Computer Vision Center and UT Austin to create a platform.” made statements.

A two-stage development program will be carried out within a 48-month period. In the first phase, Intel will focus on building new simulation platforms and mapping tools that mimic complex terrain environments with the highest accuracy.

In the second phase, Intel, which will work with RACER collaborators, will accelerate the R&D process by applying new algorithms without using a physical robot. Refinement of sim2real techniques, which is the concept of training the robot in simulation to gain skills and then transferring these skills to a relevant real robotic system, will also be carried out at this stage.

Intel says that new simulation tools that it thinks will significantly improve the development of autonomous systems will reduce the risks, costs and delays typically associated with testing and validation protocols.