DARPA supports virtual worlds for autonomous off-road vehicles • The Register

Intel has shed light on its participation in a DARPA program set up to help develop autonomous combat vehicles capable of off-roading.

The x86 Giant Tuesday describe its involvement in the US government agency’s Robotic Autonomy in Complex Environments with Resiliency – Simulation (RACER-Sim) project.

RACER-Sim is part of DARPA expanded RACER program to further the advancement of autonomous machines capable of following human-controlled vehicles over difficult terrain amid conflict and other real-world situations. RACER-Sim, as the name suggests, involves the creation of simulations in which these autonomous systems can be developed and tested before being tested in the real world. This is useful because it’s a good idea to perfect the code as much as possible in a virtual world, where it can’t do any harm or real harm, before putting it behind the wheel of expensive and dangerous hardware.

In January, DARPA said that Duality Robotics and Intel had been chosen to be the prime contractors for RACER-Sim. “The RACER-Sim portion of the program aims to extend current simulation capabilities and physics-based models to support off-road range testing in virtual environments,” said Stuart Young, RACER program manager. “These simulation environments will allow teams to test and validate parts of their autonomy stacks without having to spend a lot of time and money on field testing.”

Duality said it will work with Neya Systems, a division of Applied Research Associates, on his contributions to RACER-Sim, which includes a “20 km x 4 km photorealistic and adjustable virtual environment to simulate the various operating conditions of the vehicle”.

This week, Intel said as part of its participation in the study that it is collaborating with the Computer Vision Center in Barcelona, ​​Spain, and the University of Texas at Austin in the United States. DARPA budget for the first phase, at least, for RACER-Sim was pegged at around $6 million; Intel acknowledged that “this research is funded in part by the US government.”

Intel's RACER-sim digital environment

Intel’s rendering of a rugged buggy-like autonomous vehicle exploring one of its RACER-Sim virtual worlds

Intel said there are many simulation environments that mimic real-world conditions, “but few are optimized for the development of large-scale, high-speed off-road autonomy,” which this project aims to provide. . The conditions a vehicle encounters in difficult terrain, Intel said, are simply too great to do off-road simulations except expensive and slow…although probably not too expensive or taxing for an organization like Intel.

The chip-factory giant said its approach to RACER-Sim would take place over 48 months in two phases. The first phase of Team Intel’s plan “is to create new simulation platforms and map generation tools that mimic complex off-road environments with the utmost precision,” including physics and sensor modeling of vehicles. In particular, Intel wants to create a simulation area at “never seen before” scales of over 100,000 square miles.

Once Intel and its partners clear this first resource-intensive hurdle, the second phase will be to pair this Star Trek Holodeck-like technology with the broader contributors to the RACER program so that their robot vehicles can not only be tested and validated, but drawn into virtual worlds. .

“Intel expects these new simulation tools to significantly improve the development of autonomous systems using virtual testing, reducing the risks, costs, and delays associated with traditional test and verification protocols,” said the flea giant.

Self-driving technology is an area that DARPA has been funding for years. In 2004, the research arm of the United States Department of Defense organized its first Grand Challenge, with a million dollar prize for an autonomous vehicle capable of covering a course of 150 miles (240 km) in the Mojave Desert. without a human taking control. The results were hilarious, with none of the contestants covering more than a few miles.

The following year, DARPA reissued the challenge, this time with a $2 million prize; a team from Stanford University won it in a Volkswagen Touareg R5 equipped with seven Pentium-M processors in a rack to do the math. Intel showcased the car at its Developer Forum that year, nearly suffocating conference members by driving a diesel vehicle through the San Francisco auditorium.

Intel persisted in the field by buying Mobileye in 2017 for $15 billion. Last year, processor giant Xeon announced it would take Mobileye public to get its money back, while retaining control of the company. ®

Comments are closed.