Nascar used robotics to crash next-generation car

Nascar used robotics to crash next-generation car

Near the end of regular season for NASCAR’s Cup Series there are many takeaways: a near-record number of first-time winners, better parity and memorable races. Much of it can be attributed to NASC
R’s new Next Gen car that has performed flawlessly for the most part.

All of this has given NASCAR confidence in the new car.

Delayed by a season due to Covid, NASCAR had plenty of time to test and retest the new machine, which features many updated technology and safety improvements. And those safety improvements have been shown very publicly thanks to several hard crashes, most notably at Daytona, Auto Club Speedway, Atlanta and Charlotte.

NASCAR was also very confident in those safety innovations when they tested them as well. And that testing used new, innovative ways that were unimaginable not so long ago.

NASCAR partnered with AB Dynamics to autonomously crash its Next Gen vehicle using robots.

Last October during a test at Talladega Superspeedway, AB Dynamics outfitted a Next Gen car with electric motors, essentially robots, and a crash test dummy then programmed it to crash into a wall at 130 mph.

“The challenge was to get this highly complex machine to perform a very accurate test without a human driver operating the car,” said Craig Hoyt, Business Development Manager at AB Dynamics. “A major hurdle that NASCAR faced was finding a crash testing facility that could conduct such high-speed crash tests. AB Dynamics robots allowed NASCAR to use a fully running race car and run the test on a real track, at real race speeds. There is no better data than replicating crash tests in a real environment and our robots allow us to do that accurately and repeatedly.”

But this wasn’t a remote-controlled car, but using what Hoyt calls “serious GPS connectivity,” a pre-programmed route was used to achieve the level of accuracy and precision NASCAR needed to get within a predetermined set of parameters. . At Talladega, those parameters meant that the car had to go into the outside wall at 130 mph from the pit lane.

AB Dynamics was confident that their robotics would perform flawlessly; they have a lot of experience with robotic vehicles.

“Our typical customers are OEMs or testing bodies, such as NHT
SA (National Highway Traffic Safety Administration),” said Hoyt. “OEMs will certainly use this technology for durability testing, accident avoidance system testing and even occasional crash testing.”

Cars are no longer just strapped to a sled and pushed with their heads against a wall. With robotics, engineers can get more accurate parameters and better test results.

The NASCAR crash test in Talladega last October was actually a culmination of testing. Many of those tests involved artificial intelligence (AI) teaching the robots how to drive a race car.

“We had a lot,” Hoyt said. “When we installed the robot system in the car, we had to teach the robot to drive. By learning I mean we do some tests where the robot learns the different characterizations of the car; how it accelerates, how fast things you would naturally do when you drive a car, right? Similar to PID tuning: how much rattle is too much, when do you have to shift that sort of thing.

“A NASCAR vehicle isn’t just jumping into your car, it’s a precision vehicle; those guys are pros and it’s a precision vehicle. So we had countless tests leading up to the crash.”

Using AI to control a race car raises the obvious question: can a ‘race’ be held with only robots behind the wheel?

“That’s a good question,” Hoyt said, laughing. “Yes absolutely. Our typical use case with these robots is that we’ll be driving target robots, we’ll have vehicles being tested, and we’ll have human drivers all interacting on a proving ground.

“That could be a dozen or more different vehicles, all performing in choreographed movements. That’s the core of our business.”

However, for the Next Gen car tests, there were no multi-car tests.

“I think the primary purpose of this test was to verify their (NASCAR’s) simulation efforts,” Hoyt said, adding that they wanted to test within a limited set of parameters, “as they were trying to match a simulated effort and then verify that in the real world.”

Hoyt spent nearly 30 years in racing in CART and IndyCar. He moved to Charlotte, North Carolina about 20 years ago and is still a fan of all motor sports.

“It’s been a love of mine since I was little,” he said.

And like NASCAR’s engineers, when he sees a crash during a race with the Next Gen car, he’s convinced the driver will run away. This is largely due to testing by AB Dynamics, robots and dedicated people like Craig Hoyt who ensure that racing is as safe as possible.

“You know, that’s a source of pride,” he said. “That’s ingrained in everyone who does racing; I think that’s everyone from the sticker man all the way to the crew chief, to NASCAR itself, that’s what everyone is on about. I am happy to be a part of that.”