It’s not often you get to geek out with a Formula 1 IT director, but that’s exactly what happened recently when I spoke with Michael Taylor, IT Director at Mercedes-AMG Petronas Formula One Team during the Spanish Grand Prix in Barcelona.

The challenges facing Formula 1 teams such as Mercedes-AMG are familiar to most of us that deal with technology on a day-to-day basis – navigating the benefits and pitfalls of cutting-edge tech, AI, and balancing budgets under the sport’s strict spending caps.

Upgrade Cycles

We all get new devices on a regular basis, whether they’re smartphones, laptops, or other gadgets. But how often does a Formula 1 team need to replace its technology? Interestingly, while keeping it up to date is important to ensure it remains competitive, Mr. Taylor cited a three-year refresh cycle to deal with the physical demands on track-side hardware.

Those demands include traveling around the world plus high humidity and temperatures at certain tracks. It’s particularly hard for the drivers, which Formula1.com describes as being only fractionally cooler than your average sauna in the actual racing cars, with drivers losing close to seven pounds in sweat. This also takes a toll on the technology, which is why much of it needs to be replaced every three years due to failure rates, as opposed to becoming obsolete.

Mr. Taylor did expand on hardware upgrades that are beneficial in terms of performance.

“A new SKU of processor from AMD, for example, could have a significant benefit in terms of cycle time reduction, which is a key area of performance development. But we’re obviously operating under a spending cap; we don’t have money to burn. It’s about balancing the benefits.”

Spending Caps and Technology Limitations

In 2021, the FIA introduced financial regulations aimed at leveling the playing field across the ten teams on the grid. Prior to that, larger teams could generate more revenue and spend more on development. This effectively created a cost cap including activities such as the design, manufacture, test, and racing of the cars. Technology falls squarely within that cap, and various sections of Formula 1 teams compete for that team budget too.

Interestingly, computational fluid dynamics (CFD), which is used to design the car’s aerodynamics and effectively create virtual wind tunnels, is not allowed to be processed on GPUs and is limited to CPUs. Using GPUs would improve processing times, but there are also limits placed on the processing horsepower measured in teraflops that teams are able to use.

Aerodynamics Always Needs More Computing Horsepower

It comes as no surprise, then, that when asked which area he’d pick if allowed extra computing horsepower, Mr. Taylor said it would be CFD.

“If we were able to increase or even have unlimited resources, CFD is definitely an area we’d invest heavily in terms of refreshing equipment regularly. Obviously, we’d love to use GPUs from the likes of AMD, but currently, the FIA doesn’t allow that. If they did, it would bring a five or even ten-time improvement to processing performance.”

The importance of aerodynamics has been made clear by none other than the Mercedes team itself. Its cars have struggled to keep up with their Red Bull counterparts since the latter took the drivers’ title in 2021 with Max Verstappen going on to win it in 2022, 2023, and looks set to win this year too. However, aerodynamic upgrades brought to both Mercedes cars closed the gap in recent races, with Lewis Hamilton and teammate George Russell claiming third and fourth spots in qualifying with Hamilton also on the podium on race day – a rare occurrence since Red Bull’s recent dominance began.

If There Was One Area That They Could Completely Automate Using AI, What Would It Be?

This year the buzzword from so many companies is AI. It’s coming to laptops and software, and investment in it is huge. Formula 1 would likely see huge benefits from it, but which areas would benefit the most in a sport like this? Unsurprisingly, Mercedes’ response was that it’s an area where humans have often struggled – analyzing data, especially in real-time.

Formula 1 generates staggering amounts of data, both during testing and a race. A single car in a single race can generate 1TB of raw data, with this increasing during analysis afterwards, from over 250 sensors on board the car according to this article from Mercedes. Similarly, making sense of that data and working it into strategies for the race is also vital.

Teams such as Ferrari have struggled with strategy in recent years with The Independent’s classic quote being “Ferrari’s strategies are so bad, it’s almost like they don’t want their drivers to win”. This is undoubtedly where AI could help. In fact, using it both during and outside of races could be massively beneficial.

According to Mercedes’ Michael Taylor, AI could deliver better data-driven decisions than relying on human eyes by analyzing more data faster and more intelligently. Arguably, though, this could remove some of the dynamic chance and variation in races too, and the human factor in strategy is entertaining and a huge talking point in post-race interviews.

What Technology Has Seen Big Benefits to Formula 1 in the Past?

Technology is obviously a hugely important tool in Formula 1, which is used in practically every area. But what advances have mattered the most in the past? To Mercedes, it’s processor power and the ability to store more and more data.

Mercedes’ Michael Taylor told me “The ability to generate much larger data sets and higher fidelity models – it’s always been available to us, but the evolution of compute and year on year have 10-15 percent improvements for relatively little investment and the cost of storage coming down – those two things combined really created this data-based explosion in our world.”

I’d like to thank Michael Taylor for his time, and it will be fascinating to see what future technologies play a part in the immediate future of Formula 1 and how, especially AI, but it’s certain that technology will play an increasingly important role both during and outside the race calendar.

• 0 Comments
• Ai Process
• Artificial Intelligence