The Abu Dhabi-backed A2RL Autonomous Racing League plans to pit ten driverless cars in wheel-to-wheel battle from next year; the most ambitious AI motor sport plan yet. Its backer reveals more about the super-sized science experiment.
This weekend’s Abu Dhabi Grand Prix sees the Yas Marina Circuit host the most advanced racing cars in the world preparing to do wheel-to-wheel battle.
But although the race is on the 2024 F1 schedule, the same may not be true next year. Because, next spring, ten entirely different teams are set to compete on the track with cars that could go far beyond F1 in pushing the boundaries of science and motor sport — without a single driver.
On April 24, 2024 the lights are due to go out on the largest championship yet for driverless racing cars. Ten Dallara-made Super Formula SF23s, based on Japanese Super Formula cars, will be piloted solely by artificial intelligence (AI), each trained to compete wheel-to-wheel at over 180mph by some of the brightest software engineers in the world.
We’ve seen autonomous racing cars before, and even a handful of ‘races’, but the Abu Dhabi Driverless Racing League (A2RL), takes driverless competition to an ambitious new level, even though spectators have so far failed to warm to the idea.
“As of now, we know it’s possible to race a car autonomously on its own at serious speeds around a racetrack — we’ve done it and others have done it,” said Thomas McCarthy, executive director of Abu Dhabi state investor Aspire, which set up the series. “It’s also possible to race two cars around the racetrack where they actually engage and pass each other. But nobody has done it with more than two cars as of now.”
If all goes to plan, April will see ten cars, powered by 2-litre, four-cylinder, turbocharged engines and programmed by teams from universities around the world, light up their rear tyres and rocket towards the 90-degree right-hand Turn 1 at Yas Marina while jostling for position, feinting overtakes and taking aggressive lines, just as F1’s human pilots will this weekend.
The split-second decisions will be taken by an AI module, which will use sensors including radar and lidar (laser) scanners, as well as cameras and GPS to build up a picture of the circuit and competitors, then decide on the course of action that is most likely to win the race, using a drive-by-wire system to activate the accelerator, brakes and steering and gears.
Race weekends, initially planned to take place only at Yas Marina, will consist of sprint races, time trials and feature races, with full details still to be announced.
But A2RL has its work cut out merely to get the series off the ground, let alone race with ten cars on the same grid. It says that some teams are behind schedule in preparing for the first race, while the experience of other series with similar ambitions sounds a warning note.
Roborace’s warning shot
The Roborace project was shut down in 2022 after years of challenging development, marked by car that drove straight into a wall and another that crashed during a demonstration in front of spectators at a Formula E race. Roborace eventually ran an initial championship, also with university teams, but the cars never achieved the ability to battle wheel-to-wheel. In some races, they had to overtake virtual competitors, displayed as graphics onscreen. In others, they duelled with a rival car. If one closed to within a certain gap, the leading car would slow down, and let the trailing car overtake.
In contrast, the US-based Indy Autonomous Challenge is still going strong, but is ramping up at a slower pace. This year’s challenge brought together competing university teams that had developed their own autonomous software. Each then tested their work in an adapted IndyCar by passing another car driving at a steady speed on track. The speeds were gradually increased until just one was left standing.
A “Next Gen” autonomous car is due to be unveiled in January at the CES technology show, when it’s promised that nine teams will take part in a race at the Las Vegas Motor Speedway.
Although there’s a novelty factor, autonomous racing has not attracted large numbers of fans so far, and even the ambitions of Abu Dhabi’s A2RL series don’t extend to winning over followers of other racing series that rely on the natural instinct and adrenaline-fuelled pace of a human driver. It’s not meant for them though. This is a super-sized, motor sport-themed science experiment.
“We want it to be a sport, so that there is a certain degree of enjoyment,” McCarthy told Motor Sport. “But also potentially to be pushing ahead, with technological developments that can have near and long term benefits for humanity — that’s a real big driver for us.”
Those wider ambitions appear to be why Aspire, which invests in technology research on behalf of the Abu Dhabi government, created A2RL. The racing in Abu Dhabi will be a chance to test the latest in AI development, which could later be applied into general problem solving as well as the driverless technology that is becoming ever more prominent in our road cars. As McCarthy points out, technology that can safely race wheel-to-wheel at over 180mph, should be able to handle driving down a lane of a British motorway.
The recent buzz around AI has been driven by the release of ChatGPT, which allows anybody to interact with its chatbot. Its ‘knowledge’ comes from a machine learning model, which has scanned millions of web pages in search of language patterns until it can — with mixed results — produce passable responses to questions and requests.
Machine learning is also the basis for A2RL. Each competing team will get an initial version of the software, which was firstly trained on telemetry from a human driver. Sensor, input and GPS data was fed in to the model, allowing it to link the driver’s actions with the track layout and go on to identify competitive racing lines, braking zones, as well as more complex aspects such as the level of grip available and the effect of tyre wear.
Honing that model will now be the task of each team. The cars are — as far as possible — identical, with exactly the same sensors, so the only way to gain an advantage is to boost the AI’s brainpower to make it a more effective racing machine.
“It’s up to each team to scale the concept,” added McCarthy “In contrast, when this was done by the Autonomous Indy Challenge, they weren’t given a baseline level of software and many teams never made it out the garage. When we give them a car, they know it actually can perform and can then use it as a building block to achieve higher and higher speeds over a two week practice period ahead of the actual race.”
During the months leading up to the race, teams will be focused on building up the “stored knowledge” within the AI through simulator sessions and on-track tests to improve its decision-making in the heat of a race.
“Think of a corner on a track,” says McCarthy. “There’s of course an optimal line that the car will be programmed to follow, but then the next questions are: ‘What’s what’s the optimal speed I can come in and do that at?’ ‘How is that connected to the performance of the car itself?’ ‘How is it connected to the performance of the tyres?’ ‘How is it connected to the heat of the tyres at that point in time?’ ‘And how is it connected to the way the gear ratios are set up on the car?’ All of that data is also going into the car and it is then up to the coders to write software that can answer each question quicker than the rest.
“Essentially we’re taking the driver out the car and putting the backroom staff you’d usually find in an F1 team back in.”
Come race day, teams will set their updated software loose and monitor how it performs against rivals. Certain parameters such gear ratio and engine control settings will be locked, restricting set-up possibilities so that the performance of the car will almost entirely be down to the effectiveness of the AI each team has developed.
“Ultimately we want the coding in one of the cars to outsmart the coding in the other one,” says McCarthy. “So that it can take an optimum [racing] line and use the SF23’s acceleration to pass. Then, rationally, the other car has to recognise that.”
The A2RL is also not only an opportunity to accelerate the development of autonomous technology but also provides a chance to champion engineers that typically go unnoticed in motor racing paddocks. Without drivers to decorate, the Abu Dhabi podium will be filled by the software superstars.
“If you think about all the data that’s getting processed — of course the computer has the capability of taking in all this data — but we’re talking about the human mind’s capability to actually program and send the commands through code to do this,” explained McCarthy.
“[Instead of the driver], the programmer is the star. And in a sense, we want to personify them as well. So that the audience will get a sense of who they are, what their background is and what they’re doing here.”
The majority of the teams will be university-based with senior professors acting as the team principals. The races will be streamed online through an app as well as on sites such as Twitch and YouTube. Each team will be responsible for its own funding and sponsorship, much like F1, but McCarthy says that there will be little comparison with racing’s top tier.
“I don’t want to put ourselves in that position,” he said. “I think there’s always a role for human endeavour, in terms of taking the risk of sitting in one of these [racing cars]. Their ingenuity and their skill set and being physically resilient enough to get through each corner: that in itself is something of beauty.
“I wouldn’t want to replace that. I think what we’re doing is creating an additional type of motor sport or a different type of extreme sport that involves autonomous systems with different types of heroes and pilots, all exhibiting different human skill.”
In fact, McCarthy believes that a driverless racing series can even humanise AI. “I think a lot of the time people are becoming more and more suspicious of science,” he says. “Every second headline you’re reading is ‘Oh my goodness, what’s AI going to do? It’s going to take this power from me!’. Whereas we’re seeing the power it gives us.
“Some people will still only watch Formula 1 and IndyCar because they’re interested in the physical skill and ingenuity of a human driver.
“But hopefully those people are equally enthused or interested by the knowledge, skill and ability of a coder,” he adds. “They may not, and we may have to wait. We’re definitely not the next Formula 1. We’re something quite different.”