F1 teams harnessing AI for speed and strategy
Formula One has a generative AI-designed trophy at the Canadian Grand Prix but, novelty and sponsorship aside, the importance of artificial intelligence for speed and strategy is growing stronger all the time.
The various software systems are helping cars go faster and teams become more competitive, crunching terabytes of data to analyse simulations and cut the time it takes to get parts to the racetrack.
"The fight we are fighting every day is of hundreds and thousandths of a second," Peter Bayer, chief executive of Red Bull-owned RB, said at a recent event with software partner Epicor at the team's Faenza factory.
"For a human being it sometimes can be a bit overwhelming actually to fight for that last thousandth or milli-milli-second.
"If you take it into perspective, we fight for seven, eight, nine, 10 thousandths of a second on a track that is five or six km long and we go 350kph and ending up on a margin of 0.01 per cent. "
AI-based software like Epicor's can detect patterns in competitors' behaviour, save time and resources through automation, and do millions of calculations in real time during races.
"If you don’t know what you are doing, because we are testing, you’ll spend a lot of resources; Carbon fibre, energy, people," said RB's head of vehicle performance Guillaume Dezoteux.
"So AI can tell you 'look at that' and 'don’t look at this'. In a specific example it means you don’t need to run 100 simulations."
PHYSICAL SENSORS
It could mean replacing physical sensors on cars, during qualifying and races, with virtual ones to save weight and remove the cost of crash damage.
"I’m convinced that on AI, Formula One will be leading with use cases which today nobody is thinking of," said Bayer.
All teams operate under a $145