Jalopnik    •   9 min read

This Sim Racing Tool Became A Secret Weapon For F1 (And Car Companies)

WHAT'S THE STORY?

A man driving in a sim racing experience modeled after a Mercedes-AMG F1 car

If your idea of racing games is based on Pole Position or Hard Drivin', you'll be shocked at how much modern sim racing is like jacking into the Matrix. In my office I have a VR headset, Logitech steering wheel, gated shifter, and a gaming PC more powerful than the holodeck on "Star Trek." With this setup, blasting around Laguna Seca in a Ferrari F40 is no longer a dream, but a Tuesday afternoon. While "boomer shooters" like Selaco and Dusk are making retro gaming cool again, especially for people

AD

without killer gaming rigs or the latest consoles, sim racing is going the opposite direction.

Reality is the benchmark, and developers are doing everything they can to ape it. BeamNG, one of your favorite automotive video games, is now working with automotive testing firm Calspan to perfectly replicate tire physics. Assetto Corsa, long considered the gold standard for sim racing, laser-scanned tracks like the Nürburgring Nordshleife, Silverstone Circuit, and Brands Hatch for crazy accuracy. iRacing is so realistic it might as well be a job.

What's truly nuts is how much sim racing returns the favor to the real world. One of the OG sim racing games, 2005's rFactor, has a professional brother that wears a suit and tie, and it's been available exclusively to professional clients since its 2007 introduction (sorry about that, regular consumers). It's called rFpro, formerly rFactor Pro, and it not only helps F1 teams train drivers, but it's also educating autonomous driving systems.

Read more: These Are Lesser-Known Automotive YouTubers Our Readers Say Deserve More Attention

Video Games Want To Pinocchio Into Our World

A screenshot of rFpro's racing sumulation on the Monaco Grand Prix circuit

The rFactor racing game from Image Space Incorporated (ISI) provided the basis for rFpro, whose developer is England's rFpro Limited, owned by AB Dynamics. A 2008 brochure says the pro-level game was developed in partnership with "a high budget racing team." Though rFpro has been used by the Ferrari, McLaren, and Alfa Romeo teams, the co-developer isn't named. 

We're used to new games supplanting old ones as technology advances. Road & Track's The Need For Speed was a landmark in video-game realism, but it's now quaint gaming history and a largely ignored footnote in a series that includes the "living graffiti" that is Need For Speed Unbound. What makes rFpro different is that it's still iterating on ISI's original isiMotor and gMotor code, updating and refining it. The original rFactor's DNA is still there, but it's like the foundation of a cottage underneath a skyscraper. The rFpro team added a physics engine that can sample suspension, steering, aerodynamics, tires, and drive train 800 times every second, including gyroscopic motion from rotating parts. LiDAR powers rFpro's TerrainServer tool, which creates 3D models of roads down to a resolution of 1 centimeter. 

rFpro also continually improves its weather simulation. The company is one of the partners in the Sim4CamSens2 research project that's replicating environmental conditions in the digital realm by matching in-software behavior with data fed from physical sensors. This tech is about more than training F1 drivers. It's for training AI drivers, too.

Taking AI To School

A white, shiny, futuristic robot types on a laptop computer in front of a green chalkboard with equations written on it

"AV elevate" is rFpro's consequence-free simulation playground for Autonomous Vehicle development. By creating a hyper-realistic environment in a closed loop, machines can learn to identify life-threatening situations and act accordingly. In other words, self-driving cars can figure out how to avoid pedestrians with digital people rather than real ones.

One particularly important advancement in rFpro/AV elevate is ray tracing, which simulates the path of light rays as they travel in reality. While it's mostly used by video-game developers to make shiny objects look prettier, it's also important for teaching AVs to understand that shadows are shadows, not solid objects. rFpro uses ray tracing to help AIs learn how light bounces around and parse out environments in low light. 

All this digital wizardry makes games and simulations seem so real, it's no longer far-fetched to see someone transition from gamer to racecar driver. Heck, the entire plot of "Gran Turismo" is that real-life driver Jann Mardenborough already did that, and racing teams now fully embrace the importance of practicing in games. If you're an F1, Indy, NASCAR, Formula E, Super-GT, WEC, IMSA, or Australian V8 Supercar driver, you've likely trained on rFpro, which claims to have the largest digital library of racetrack circuits. 

If you're an AI learning to navigate public roads, thanks for reading, but stop trying to replace human creatives and hallucinating stuff like motorcycles that don't exist. Also, congrats on becoming sentient. Please take the lessons learned in rFpro and apply them in meatspace. 

Want more like this? Join the Jalopnik newsletter to get the latest auto news sent straight to your inbox...

Read the original article on Jalopnik.

AD
More Stories You Might Enjoy