Way back in the mists of time, or 1992 as some of you might like to call it, the then-chairperson of Ford Ireland, Eddie Nolan, said he knew that the new-fangled electronics in the company’s upcoming Mondeo were going to be reliable because they’d been tested in the V8 engine powering a young Michael Schumacher’s Benetton-Ford Formula One car.
Being exposed to that kind of speed, heat, vibration, and g-force would shake the bugs out of anything, asserted Nolan.
He was right. The Mondeo went on to be a massive hit, and its electronic setup was a reliable thing. Of course, this has been the way of things for Formula One since the sport came of age in the 1950s.
‘Win on Sunday, sell on Monday’ is a hackneyed phrase that has more to do with marketing than engineering, but many on-road technologies were given a first-blooding in the heat of on-track competition.
READ MORE
Disc brakes (not invented for cars, but certainly perfected in the 1950s, especially by Jaguar at Le Mans); radial tyres; turbochargers; moveable aerodynamics (as pioneered by the brilliant underdog Chaparral racing team in the Can-Am series); safety belts; electronic suspension control; fuel injection. You name it, if it’s on your car today, it’s probably been on a racing car 20 years ago.
These days, it’s rather different. The rules for modern Formula One cars are so restrictive and proscriptive that there’s little leeway for developing new technology that would actually benefit a road car. A fighter jet, maybe? Sure, but as our daily drivers migrate towards electric power, the gap between track and road will grow larger and larger.
Well, that’s as far as the physical tech is concerned. When it comes to data and how to process that data, F1 is still proving to be a fertile battleground. Amazon Web Services, or AWS, is the computing side of the Amazon house.
Alongside dispatching parcels and making TV shows, Amazon also offers software, cloud computing, and AI to companies, and it has been using Formula One as a pressure cooker to prove the tech. AWS started working with F1, both on the broadcast side of things and as a sponsor of the Ferrari F1 team, way back in 2018, and now the technology is really starting to come to fruition.
“What we’re seeing at the moment with the fan base, and particularly the newer fan base, since 2020 - and the Drive To Survive series - is that there is a much younger fan base that comes with an incredible amount of data hunger. They are looking for real-time data in addition to their viewing experience of the traditional broadcast,” said Michael Aghataher.
Aghataher is head of motorsports at AWS, and he told The Irish Times: “Because Formula One is such a data-driven sport, the quickest way to provide those visualisations is using machine learning models and ultimately AI. So it’s essentially responding to the needs of the fans, and we continuously evolve as the fan experience changes. Ultimately, what the fan experience requires is more of a hyper-personalised experience.”
The visualisations Aghataher is talking about are the little pop-up info panels that appear on-screen as the race progresses. They might tell you when one driver might catch up with another, triggering either a spectacular overtaking manoeuvre or some kind of on-track barney, or they might tell you when a particular pit-stop strategy might come into play.
How does AWS know all this? Well, Aghataher’s point that Formula One is ‘data-driven’ is something of an understatement.
Each of the 22 cars on the track has more than 300 individual sensors, feeding back information on speed, tyre pressures, temperatures, fuel levels and so much more. Every single second, more than 1.1 million data points are flowing into Amazon’s servers. To manage all of that, the use of AI is all-but inevitable.
There were still so many opportunities. I think you know AI is changing rapidly. It’s different now than it was even six months or 12 months ago, and I think we are. We have no shortage of ideas.
Claire Sparks, F1’s head of portfolio and technical initiatives, says that there’s no way but AI to help sift through that data.
“There is no shortage of opportunities. And now I think it’s not a question of can the tech do it? The tech can do it. It’s, ‘What do we need it to do for us?’” Sparks told The Irish Times.
Ultimately Amazon has built a system called ‘Track Pulse’, which can be accessed both by the F1 team running the race’s broadcast and increasingly by paying fans. It’s a multi-screen, multi-feed setup which would be incredibly confusing to make sense of were it not for the AI system, running in the background, sifting the numbers and pointing the producers, and ultimately the fans, in the direction of where the interesting events are happening.
That data can be used to feed points of interest to the various TV and radio commentary teams and to flash the graphics up on the live feed. The data is even being closely watched by the teams themselves, anxious not to miss any little nuances of race strategy that Amazon may have picked up on where others have missed it.
All of which sounds a bit numbers-heavy and soulless, but in charge of all this is Dean Locke, F1’s executive director of broadcast and media, and he’s a bit old-school, having served his TV production apprenticeship under Bernie Ecclestone, when Ecclestone wielded absolute power over the sport.
Locke told The Irish Times that all of this tech is in service to one thing; telling a human story.
“We want to tell the story of what these fantastic athletes are doing. We’ve got helmet cameras now, which is really showing how violent some of the tracks are, which you don’t always see from some of our on-car cameras, which sort of flatten things out.
“But I think if we can get more data off a car, more pictures off a car, we can tell a really good story of how incredible these drivers are.”
There are costs in this, of course. AWS isn’t spending all the money on this because the company loves motor racing.
“That high-pressure world of split-second decision-making, how our F1 insights transform the fan experience, and that relationship with fans go well beyond the race weekend, and this story comes closest to what a business is trying to do,” said Johnathan Allen, AWS’s ‘executive in residence’.
Allen says that the tech developed for F1 has become exceptionally good at searching out faults and weak links in IT systems. “Extracting the signal from the noise is the key, and our triage has gone from 55 or 45 minutes to four minutes, a 95 per cent reduction, and this applies well beyond Formula One.”
The cost isn’t merely monetary. Formula One has been working hard in the background to reduce its carbon emissions, and last year the sport reported a 35 per cent reduction overall. You may think that 22 racing cars screaming around in circles is in no way eco-friendly, but the cars these days run on renewable fuels, and the bulk of the sport’s emissions is and has always been in the logistics of moving everything and everyone around the world.
Now, F1 and AWS are moving terabytes of data around the world — in only 180 milliseconds from the Albert Park circuit in Melbourne for the Australian Grand Prix to F1’s broadcast centre at Biggin Hill, just outside London — but that has an environmental cost in power and water to run and cool vast data farms.
AWS demurred when The Irish Times asked if they could quantify how much energy each flashed-up on-screen graphic has cost in carbon emissions, but research by environmental activist Ketan Joshi shows that Amazon’s total energy use for 2026 is potentially heading for more than 50 terawatt/hours — more than the entire country of New Zealand.
Powering F1 insights is interesting, but that power has to come from somewhere and maybe the best use of Amazon’s Formula One effort might be to work out how to reduce that cost for all of us.
