Last Updated on November 25 2023, 3:23 pm

In a significant step towards enhancing precision in Formula 1 race monitoring, the FIA has announced a new initiative to incorporate artificial intelligence for better tracking of track limits. This cutting-edge technology is slated for a trial run at the Abu Dhabi Grand Prix and if take over a significant duty currently performed by race stewards.

The FIA is set to integrate a “computer vision” program that performs shape analysis on video images to accurately identify when cars breach track boundaries. This innovative approach is expected to dramatically cut down the instances requiring manual review, thus expediting the entire adjudication process.

 

The goal is for the AI system to lighten the load on the FIA’s Remote Operations Center (ROC) by filtering out clear-cut cases, enabling quicker resolution of track limit infringements.

Tim Malyon, the head of the ROC, highlighted that the primary purpose of implementing AI is to sift through the numerous checks currently done by hand:

“At the moment we have ‘brute force’ the situation, by saying, ‘We have to do thousands of checks, how do we do that?'”

“We’re putting people on it because it’s the most accurate solution. What we’re trying to do now is introduce a level above ROC, and that’s AI software.”

 

FIA emergency rule to crack down on Verstappen antics

 

Drawing parallels with medical diagnostics, Malyon explained that just as computer vision aids in eliminating clear non-cases in cancer screenings, similarly, AI in F1 would dismiss the obvious non-infringements, allowing experts to focus on more ambiguous instances.

“Again, it may sound strange, but the methodology of this AI has many parallels to the discussions currently taking place in medicine, for example about the use of computer vision to scan data from cancer screening,” he says.

“They came to the conclusion that they don’t want to use computer vision to diagnose cancer, but rather to screen out the 80 percent of cases that are clearly not cancer, so that the well-trained people have more time to study “Look at the 20 percent. And that’s exactly our goal.”

At a typical Grand Prix, the ROC deals with around 800 checks. The hope is that AI can reduce this to just 50, making it a far more manageable task for FIA personnel.

 

Gunther Steiner to feature in comedy series

 

In addition to AI, the FIA is also testing a new precise vehicle tracking system known as Catapult, which employs receivers for detailed positioning, akin to technologies used in the NFL.

Chris Bentley, the FIA’s head of information systems, explained that the Catapult system could enhance live broadcasts, allowing for accurate “lines of interest” to be drawn and the AI to refine its assessment capabilities over time.

“There are examples from the NFL where every player on the field can be identified, even if they’re in a big crowd,” he says.

“We can also use this technology for our live broadcasts. This will be the same as the new tool and we will be able to draw the ‘lines of interest’. And then the AI ​​will learn over time.”

Malyon adds: “What we’re trying to do going forward is to improve all these technologies and introduce new ones.”

“Vehicle tracking will continue to be developed to improve accuracy. We also plan to double the number of ROC staff from four to eight next year and double the bandwidth of the connection between the circuit and Geneva to bring even more staff from the Being able to work remotely.”

 

 

Wolff and Vasseur summoned to stewards

 

 

Malyon revealed that despite the advances in technology, the FIA found that the most reliable method for monitoring track limits, particularly following the issues encountered in Austria, was the human eye. It outperformed vehicle recognition systems and timing loop data.

As a result, the FIA has deactivated the timing loops for track limits enforcement across most circuits, relying instead on video analysis by data analysts. Bentley emphasized that if a breach is too close to call, the benefit of the doubt should be given to the driver, underscoring the continued importance of human judgment in the sport.

 

 

FIA current procedure criticised by F1 drivers

 

 

“We concluded that the time loops were not accurate enough and that by far the most accurate solution was to have a data analyst watch the video themselves,” he says.

“That’s actually an interesting aspect of the story because right now it’s still humans who are ahead in loop positioning, GPS positioning and so on.”

Bentley added: “We have now switched off the loops for every track unless there is a chicane because it gets in the way of what we want to achieve. And ultimately the rule of thumb is: if it’s too close, you decide in case of doubt for the driver.”

 

 

Looking ahead, the FIA is dedicated to refining these technologies and incorporating new ones, intending to double the ROC staff and enhance the bandwidth for remote operations, thereby improving the overall efficiency and accuracy of Formula 1 race management.

READ MORE: Helmut Marko reveals Lewis Hamilton’s one condition for joining Red Bull

 

 

MORE F1 NEWS: Toto Wolff accuses Christian Horner of lying

The bitterness that developed between Mercedes Toto Wolff and Red Bull’s Christian Horner during the 2021 championship year shows no sign of abating now two years on. What should have been a nothing story reported in the Daily Mail about Lewis Hamilton’s team approaching Red Bull earlier this season has now descended into farce with claim and counter claim being made.

The article in the British media claimed that a ‘known representative’ of Hamilton had contacted Christian Horner while the seven times world champion was negotiating his new contract with Mercedes. Given that Red Bull look set to…READ MORE ON THIS STORY