The digital systems used to inform and protect the McLaren F1 Team and esports teams are seeing a huge boost from the use of artificial intelligence and machine learning to help take a leap forward in the competition.
From telemetry to cybersecurity, the amount of data collected in Formula 1 is massive, and understanding that complex data is often extremely important, particularly in an environment where speed is paramount.
TechRadar Pro Had the opportunity to speak to Ed Green, Chief Commercial Technology Officer at McLaren, and James Hodge, GVP and Chief Strategy Adviser to the team’s data platform provider, Splunk, about where AI fits into the equation, and how it can help protect the company’s digital world and enhance its process. Decision making in it – in addition to its limitations.
security and decisions
As you might imagine, security is important to McLaren in all of its operations. For the McLaren Shadow esports team, Green described a typical setup:
“If you have eight players on stage, that means there are eight PCs behind them, and maybe four more directing and cutting the show, and that’s how you end up with 24 PCs in common.”
To protect all of these devices, Green explained “We have standard endpoint protection that we put in all over the property. We use tools internally through various cybersecurity partners to monitor how our traffic is moving, and we have firewall providers to see exactly where the traffic is going.”
Although McLaren has been silent on the exact software they use, Darktrace has been known to be used in their security mode to certain degrees.
Cyber security must also be light to avoid depleting power from simulation platforms. “Usually a lot of things are fairly lightweight, so people don’t want a lot of factors on their devices doing the little bits,” says Green.
“We have natural endpoint customers that we use across McLaren, and they report on a whole lot of dashboards that can be useful – I can get an overview of that to monitor during the race.”
Green also explained that AI and machine learning are used for the team’s cybersecurity, not just for race data:
“We’ve used a lot of machine learning and artificial intelligence throughout [the cybersecurity] space, and in years past that meant our cyber security team would be filled with lots and lots of alumni; It’s a really tedious and tedious job to sit there and dig through lines and lines of cybersecurity information. “
“Now, by using a lot of machine learning and AI, we don’t have a big cybersecurity team, but they do have a more relevant context, so they can see where the information is going, and so the adoption of machine learning and AI is really important to us.”
“When you look at AI in cybersecurity, or in general, it’s either there to help you be more efficient, to help you integrate and solve really big complex challenges, or it’s there to provide you with additional help,” he added.
“In cybersecurity, in the race team, and in strategy in particular, AI is there as a decision aid; By giving these people the next best decision or helping them simulate what might happen, it means that when time pressure is high, we can make the right decision.”
Although AI in this context is mostly used for a real-life Formula 1 team, Green has suggested that it may play a role for an esports F1 team as well in the future.
The importance of the data
Data provider Splunk began its relationship with the McLaren Formula 1 team in 2020 as a platform to supply all-important remote data to cars, before later signing up to help support the Shadow McLaren esports team.
Hodge showed how more advanced and predictive calculations could be made using his AI tools. He mentioned the example of predicting tire degradation, which in the game can be affected by several factors such as the temperature of the virtual track and the level of aggressive driving:
“We can start doing predictive analytics to say ‘where we think we’re going to get to a point where the tires are no longer running versus coming to a pit stop,’ and that’s where we start looking at in-game telemetry to help make racing decisions.”
Hodge echoed how AI can help with decision making rather than being the decision-maker. When it comes to the involvement of AI in stopping strategy, for example, Hodge said:
“You might not want the AI to light up to say ‘hole now.'” You might want the human in the loop to say, “Actually, we couldn’t add that data feed to this model, so it’s not quite right.”
In explaining why it is difficult to automate decision-making, Hodge gives the hypothetical example of using AI to control lights in a home:
“It starts simple: When I walk in the room I want them. Well, how long do they have to stay? So you don’t see any movement, or they have to stay until midnight because I always go to bed at 11:30pm. Well, I just stayed up.” I’m late for a movie, so it’s twelve o’clock; I’m watching a movie not to move, so I turn off the lights. So in reality, what seems like a simple problem becomes very complex. Now, when you think about it in enterprise technology, it gets a lot more difficult. .”
He stressed the importance of creating enough data before relying on AI tools. And even alongside AI, traditional statistical methods of forecasting still have their place:
“I think it’s about layers upon layers [of data]. So when we look at cyber security for example, can we first monitor everything in the whole world? – This is where we started to see different security teams and IT control teams come together a lot, because they all want to monitor and contextualize everything that happens digitally. ”
“Now let’s look at statistical outliers. That’s usually a great place to start. Then we can add a little bit of basic predictive modeling associated with ML, and then, in the context of cybersecurity, look at taking a lot of different indicators together, and say, ‘Do you mean these Potential statistic compromises now that there’s a higher chance James is a bad actor?” That’s when you get more into AI.”
He also cautioned against considering practical concerns when developing AI:
“You also have to look at how far you want to push it and where is the best amount of effort to invest. Because often the stat side gets you close enough to where you want to be. You can spend a long time getting the perfect AI model, and almost a waste effort and money in doing so.”
“I’m a big believer in getting the fundamentals right, because no company in the world has the fundamentals perfect. The more you can do that, the better able you are to push the decision-making process into your front-line employees to do what they’re hired to do.”