Why Artificial Intelligence could be F1's next big thing

In a sporting series as technologically cutting-edge as Formula 1, data has become king. Making progress up the grid these days doesn't happen if there aren't hard numbers on a screen telling you that something is working. Even for a design genius such as Adrian Newey, if the computer says no: forget it.

From windtunnel runs and CFD evaluation, to driver simulators, track testing or laps over a race weekend, at every step of the way teams are logging what's going on to try to help find the answers that will bring them the maximum performance.

With terabytes of data flowing around F1 garages and factories, advantage comes not from gathering more than anyone else, but by being able to process what you have faster - and in particular prioritising the most valuable answers that can help you unlock something.

It's the endless chasing of needles in a haystack. In the fast moving world of F1, you can't afford to waste a whole day rummaging through the whole mound to find the prize: you need it right now. Waste a run in free practice without the answer, or a whole cycle of a windtunnel, and that's lap time lost.

Getting data from the cars was the first stage of this revolution. As an example of how far things have come, Williams's race-winning FW07 from 1979 was the team's first that had sensors on it - and it would take 20 minutes to download one lap of data for analysis.

Nowadays, Mercedes proudly boasts that through a partnership with Qualcomm to utilise state-of-the-art wifi transfer, it can have all the data it wants available in the pits before a car has come to a standstill between runs.

F1's data push has taken place swiftly and without much fanfare, but the changes in the pitlane have been fairly significant in moving the championship away from behaviour that had been commonplace since cars first raced.

"Just like the internet transformed our world in the last 20 years, blockchain technology, or distributed digital ledgers, will transform us again" John Zanni from Acronis

Williams's chief information officer Graeme Hackland says that even since he joined the team in 2014 he has seen a big shift in procedures becoming more digital.

"F1 is all data driven," he says. "And that means every decision that is taken for the whole lifecycle of the car, not just whether we are going to pit now or which tyres to put on.

"Digital has even taken over some of the manual process. The tyre mechanic, when I arrived in 2014, had pieces of paper and he would write information about the tyres - like temp, pressures, and which set was for which driver.

"He would write it down on paper, go back to his computer, type it all in and then the engineer on the pitwall would see that information.

"We have now digitised that process such that the engineer sitting on the pitwall can see the information live. It is now about getting data into the hands of engineers when they need it in order to make the right decision.

"In the past, the engineer would have said 'I want that set of tyres', and it may or may not have been the right set to choose, as the info he had was five minutes old. Now it is all done in real time."

One of F1's mantras is that if you stand still you go backwards. So this overriding value of data means that boundaries are needing to be constantly pushed. It's about getting more bang for your buck and finding any kind of incremental gains to get a lap-time advantage over your rivals.

That is what has made F1 so appealing for a company such as Acronis, which last week announced a tie-up with Williams in a bid to further push the barriers of what is possible.

John Zanni, Acronis's president, is always full of enthusiasm about F1 whenever you chat to him; it's an area of his business where barriers are there to be broken.

"The rate at which technology is changing is just mind-boggling," he says, during a chat at Williams last week.

But the real fascination comes not from what is happening now, it's about what's coming next.

For example, one of the big buzzwords being thrown around the business community now is 'blockchain' - and how the wider world can make use of a public ledger rather than having everything on a central database.

Essentially blockchain is a database of information not stored in a single place, and formed of blocks of data linked together in a chain - meaning there is a register of any changes.

While some companies have cheekily jumped on the bandwagon and added 'blockchain' to their name to help boost their share prices, Zanni actually sees some viable benefits for the technology being used in F1.

"In the case of F1, right away, it would solve the problem of verifying the problem that your data isn't corrupted or hasn't been changed," he explains.

"When you talk about data that is five, 10 or 15 years old, if you have put a fingerprint of that data in the blockchain, then even if we are not around anymore, you can see whether that data has been modified or not. So that is one example.

"The industry recognises that it is transformative technology. Just like the internet transformed our world in the last 20 years, blockchain technology, or distributed digital ledgers, will transform us again.

"What happens in the next 20 years, what applications come up, is hard to predict though. All that I know is that if you are not experimenting with it now, at least having something in production now, it will be hard to keep up. I've been to a lot of conferences where I've seen a lot of these start-ups, and they are all full of ideas - so it is just a question of what will take."

"By the time we get the data two days later, the engineers either no longer need it or we've not been able to add the performance needed for the next race" Williams's Graeme Hackland

But while there is little doubt blockchain will come to change our world - perhaps including F1 - over the next decade, it is not this area where Zanni thinks the hype will be during the next 12 months.

For him, it's Artificial Intelligence (AI) that will become one of the main technology battlegrounds over the next year or so.

We are not talking about clever robots sat on a pitwall using their algorithms to make the right strategy call. The AI being talked about right now will be used at first to help better manage access to data.

The computers will learn to know which data needs to be saved; which data needs to be prioritised so there can be rapid access to it. Plus it needs to be one step ahead and bring up data that is needed next.

The task won't be to find that needle in a haystack as quick as possible; it will be about ensuring that needle is in your hands before you've even thought you need it.

Zanni adds: "We are looking at how we can use deep learning and AI technology over the next year to two years, to not have to manually think about it and automatically do it. It is a big challenge."

Even for a team such as Williams that has undergone a big digital overhaul, the change from what there is now to what could be possible with AI would be dramatic.

As Hackland explains: "If someone wants something from the past three months, then I can guarantee you we can get it fairly quickly - and that means half a day to a day. If we go further back then it can take even longer. That is not good enough in an F1 context.

"By the time we get the data two days later, the engineers either no longer need it or we've not been able to add the performance needed for the next race."

Getting AI to do all this seamlessly could dramatically accelerate how quick teams can progress.

Zanni said: "The issue will be AI as a core function of specific applications, whether in F1 or in data protection, versus artificial intelligence as a kicker or boost to something you are already doing today. That will be the difference.

"We have some systems that can go through videos and look for all the videos of you, and categorise them. This is using a lot of the computing power of public clouds to be able to do this. How we turn this into something really, really useful, will probably take another six-to-12 months.

"But the fact Williams are willing to take risks and try new things helps us a lot, because it means as long as we are meeting their needs with the core tasks, like back-up and recovery, we can work with them on new market leading technologies. And that will give them a competitive advantage."

So be ready F1. The bots are coming.