The first-ever race between four self-driving cars and a Formula 1 driver just happened in Abu Dhabi

(CTN News) – Numerous computer displays with telemetry data exist in the pits of competitive racing events, such as Formula 1. Modern teams are inundated with real-time digital feedback from their automobiles.

I’ve been in many of these pits over the years and marveled at the torrents of data, but I never saw an instance of the Microsoft Visual Studio software development suite running amid the turmoil.

But then, I’d never seen anything like the inaugural Abu Dhabi Autonomous Racing League race, which took place this past weekend. The A2RL, as it is called, was not the first autonomous racing series. The Roborace series featured self-driving race cars navigating virtual obstacles, while the Indy Autonomous Challenge was held at Las Vegas Motor Speedway during CES 2024.

While the Roborace focuses on single-car time trials and the Indy Autonomous series on oval racing, A2RL aims to explore new ground in a few areas.

A2RL put four vehicles on the track for the first time and had them compete simultaneously. Perhaps more importantly, it paired the best-performing self-driving car against a human being, former Formula 1 pilot Daniil Kvyat, who raced for many teams between 2014 and 2020.

The true struggle was behind the scenes, with teams staffed by an astonishingly diverse cadre of engineers, ranging from aspiring coders to doctoral students to full-time race engineers, all competing to find the limit in a whole new way.

The A2RL race cars are standardized to ensure an equal playing field, unlike Formula 1, where 10 manufacturers create unique cars with AI assistance. The 550-horsepower machines borrowed from the Japanese Super Formula Championship are identical, and teams cannot replace any components.

That includes the sensor array, which consists of seven cameras, four radar sensors, three lidar sensors, and GPS, all used to comprehend their surroundings. While touring the pits and conversing with the various teams, I discovered that not everyone fully utilizes the 15 gigabytes of data each car generates with each lap.

Some organizations, such as Indianapolis’ Code 19, have only recently begun working on developing a self-driving car. “There are four rookie teams here,” Code 19 co-founder Oliver Wells stated. “Everyone else has been competing in competitions like this, some of them for up to seven years.”

TUM, located in Munich, and Polimove, based in Milan, have substantial experience competing in and winning Roborace and the Indy Autonomous Challenge. That expertise, along with the source code, is carried over.

“On the one hand, the code is constantly being developed and improved,” explained Simon Hoffmann, team principal at TUM. The crew adjusted the cornering behavior to suit the steeper corners on the road course and the overtaking aggression. “But generally, we use the same base software,” he remarked.

The teams with the most experience dominated the timing charts during a weekend of qualifying rounds. TUM and Polimove were the only teams to complete lap times under two minutes. Code 19’s fastest lap time was just over three minutes, but the other new teams were significantly slower.

This has resulted in unusually fierce rivalry in software development. This one is unique, unlike past competitive coding challenges like TopCoder or Google Kick Start. Improvements in code result in faster lap times and fewer crashes.

Kenna Edwards is a Code 19 assistant racing engineer and an Indiana University student. She had some previous app development knowledge but needed to learn C++ to design the team’s antilock braking system. “It saved us at least a couple of times from crashing,” she told me.

Improved algorithms produce tangible outcomes, unlike traditional code challenges that may necessitate debuggers or other monitoring tools. “It’s been fascinating to see the flat patches on the tire improve throughout the next session. “They’ve either reduced in size or frequency,” Edwards stated.

This application of theory creates interesting engineering difficulties and opens up viable employment opportunities. After interning with Chip Ganassi Racing and General Motors, Edwards will begin full-time employment at GM Motorsports this summer, thanks to her expertise with Code 19.

This type of development is key to what A2RL is all about. A supplementary series of competitions for younger students and youth organizations worldwide runs alongside the main on-track activity. Before the main A2RL event, those organizations competed with self-driving 1:8-scale model automobiles.

“The goal is to keep the smaller model cars for schools next year, and for universities, we may do it on a little larger go-karts, so they can play with the autonomous go-karts.

And then, if you want to be in the top league, you start racing on these cars,” explained Faisal Al Bannai, secretary general of Abu Dhabi’s Advanced Technology Research Council (ATRC). “I think by them seeing that path, I think you’ll encourage more guys to come into research, to come into science.”

Al Bannai’s ATRC is bearing the tab for the A2RL, which includes everything from the vehicles to the accommodation for several teams, some of which have been testing in Abu Dhabi for months. They also threw a world-class celebration for the main event, complete with concerts, drone races, and an insane fireworks display.

The on-track action was slightly less remarkable. The first attempt at a four-vehicle autonomous race was canceled because one car spun and obstructed the next cars. The second race, however, was even more dramatic, with a pass for the lead when the University of Modena’s Unimore team car ran wide. TUM made the move and won the race, taking home most of the $2.25 million prize money.

In terms of man vs. machine, Daniil Kvyat made easy work of the autonomous car, passing it not once but twice to enormous shouts from the collected throng of over 10,000 fans who took advantage of free tickets to see a little bit of history—plus approximately 600,000 more streaming the event.

The technical difficulties were unfortunate. Still, it was an amazing occasion to watch, demonstrating how far autonomy has come—and, of course, how much more progress is still to be made.

The fastest vehicle was still at least ten seconds behind Kvyat. However, it completed nice, clean laps at an astounding speed. That’s in stark contrast to the first DARPA Grand Challenge in 2004, when every racer either crashed into a barrier or wandered off into the desert on an unintended adventure.

The real test for A2RL will be whether or not it can become a financially sustainable series. Most motorsports are driven by advertising, but this one has the added benefit of developing algorithms and technology that manufacturers may plausibly use in their cars.

ATRC’s Al Bannai informed me that while the series organizers control the cars, the teams own the code and can license it: “What they compete on right now is the algorithm, the AI system that makes this car do what it does. That belongs to each team. It does not belong to us.”

The true race may not be on the track but forming alliances with manufacturers. After all, what better way to instill confidence in your autonomous technology than to demonstrate its ability to handle traffic on a race circuit at 160 mph?

Source techcrunch.com