The difference between success in Formula 1 often lies in tiny milliseconds and millimeter distances. Racers know exactly which track to choose in the corners to achieve the optimal lap time. Sometimes, however, it happens that competitors try to gain an advantage they will go beyond the borders. In order for the judges to check whether the car’s wheels have actually crossed the white boundary line, F1 will test artificial intelligence system.
The Fédération Internationale de l’Automobile (FIA) has announced that this weekend’s season finale Abu Dhabi Grand Prix will use computer vision technology. For the time being, the FIA has no plans to fully automate on-track limit infringement reviews. Rather, he wants to significantly reduce the number of potential rule violations that are sent for manual review.
As Reuters notes, at the Austrian Grand Prix in July, four people had to review around 1,200 potential violations. After several on-track limit violations went unsanctioned at October’s US Grand Prix, the stewards recognized they would have to find a new way.
Computer vision is used in medicine to reviewing cancer screening data. “They don’t want to use computer vision to diagnose cancer, they want to use it to rule out the 80 percent of cases that are clearly not cancer, so that well-trained people have more time to look at the remaining 20 percent,” said Tim Malyon, the plant’s deputy director and head of remote operations FIA with the fact that same approach also chooses on the track.
The FIA hopes it will succeed reduce the number of possible offenseswhich the committee manually reviews, to approx 50 per race. According to Malyon, the goal is to remove those that clearly don’t need human review. Although the FIA won’t rely entirely on artificial intelligence to decide races anytime soon, Malyon expects that to happen eventually.