What many F1 fans have always said from their gut has been scientifically proven: Ayrton Senna was the fastest driver of the past decades. The Brazilian, who died at the 1994 San Marino Grand Prix, is ahead of Michael Schumacher and Lewis Hamilton.
“It’s nice that we were able to confirm this gut feeling with huge amounts of data and a very sober algorithm,” says Rob Smedley in an interview with Tagesspiegel Background. The Briton has been director of data systems for Formula One Group, the racing regulator, since 2020. He previously worked as a data analyst and racing engineer for the Williams, Jordan and Ferrari teams for many years.
In his new position, the graduate mechanical engineer and mathematician works closely with Amazon Web Services (AWS). As a technology partner, it supports AWS Formula 1, which was sold to US company Liberty Media in 2016 for a relatively modest $4.4 billion. With high-performance computers, artificial intelligence, and big data, AWS helps race teams develop their cars faster and more efficiently. And fans must be able to follow the performance of teams and their drivers in real time, and any wrong decision costs how much time.
Data evaluated from 1983 to the present
Perhaps the best result of this cooperation: the fastest driver. Under this heading, IT geeks evaluated historical data from 1983 to the present for several months in order to choose the fastest engine of the modern era. In order to calculate race luck, weather and accidents, machine learning specialists compared data from qualifying laps before drivers from the same race team. Outliers that were more than 2 seconds apart, such as those from accidents, were omitted.
After gathering this data, AWS professionals created a network of peer comparisons with the goal of comparing drivers across teams, seasons, and circuits. For example, Sebastian Vettel and Max Verstappen have never been in the same team before. So driver Daniel Ricciardo acted as a link: Ricciardo was averaging 0.18 seconds per lap slower than Verstappen in the 2016 to 2018 seasons with Red Bull. But when he drove with Vettel on this team, he managed to push Hessian out by 0.1 seconds.
As a result, current world champion Verstappen finished fourth in the long-term standings – 0.280secs per lap ahead of Ayrton Senna. Vettel finished 10th with a time of 0.435 seconds.
Many world champions are behind
The fact that drivers like Fernando Alonso, Nico Rosberg and Charles Leclerc are 5-7 is no surprise to F1 fans – but Heikki Kovalainen and Jarno Trulli are still ahead of Vettel. “This was also unexpected for me,” says Smedley, chief data officer. “But if you take a closer look at the courses, you can see that they are correct. These are very pure results, without any manipulation.”
The disappointing placements of multiple heroes Alain Prost (20th), Nigel Mansell (28), Mika Heikkinen (38) and Nelson Beckett (40) show that sober characters sometimes don’t quite reflect reality. Finland’s Hakkinen, who snatched the world title from Schumacher in 1998 and 1999, is 0.588, a huge shortfall for Formula 1 conditions on top-level Senna. However, in 161 Formula 1 races, he has secured 26 first grid positions, 25 fastest race laps and 20 Grand Prix victories.
“They’re all great riders, but that doesn’t mean they’ll outgrow that hot lap on Saturday afternoon,” Smedley says. Qualifying is “raw speed only,” while racing is also largely about tactics. And there was, for example, Proust, whom they called the “Professor”, Professor.
300 sensors installed in each car
Notable drivers in Formula 1 history have also excelled in developing their own cars. Niki Lauda – who drove too early for the fastest driver rating – was praised for his “pressure gauge”. Today, pilots and teams are powered by high technology. 300 sensors are installed in each car, they send more than 1.1 million data per second to the pits.
“It gives me a competitive advantage over others,” Smedley says. The more consistently and intelligently the team uses this block of data, the faster it can develop its car. “The competition is very tough, if you don’t use the latest technology, you won’t be able to survive.”
However, the data and high-performance computers to evaluate is not just important in order to have a better car than the competition before and during the season. They were instrumental in setting regulations for the 2022 car and speeding up aerodynamic simulations by 70 percent.
Because of the lower contact pressure, there are more overtaking maneuvers this season, and the championship is more open. “Now all of a sudden Ferrari is in front, and Mercedes isn’t in front, the midfield is getting closer to the first group,” says Smedley, 48, a Formula 1 veteran.
The fans are happy. However, they should not only be linked to the Netflix series about the racing series, but also by the entirely new insights that AWS enables through its data. Liberty Media is talking about “data entertainment”.
This means: Spectators receive real-time information about teams’ strategy, pit stops, engine performance drops, and tire resolution. It is analyzed in front of their eyes to see which driver is a good start and which braking technique he has.
Late and brutal braking isn’t everything, turning into a curve, driving through a curve and accelerating is a tricky interaction that can be better understood by data and not just by classic slow motion.
Seven Key Measures of Seasonal Pilot Performance
The strengths and weaknesses of every car and driver are highlighted. From this data, AWS calculates the seasonal performance of pilots. The seven key metrics – qualifying races, race start, race first lap, race speed, tire management, and a driver’s ability to stop and overtake – are set on a scale from 0 to 10 to calculate the “score” for fans and teams.
But Formula 1 should not only remain exciting, but also become greener – Rob Smedley agrees. According to the 2026 powertrain regulations, the proportion of electric vehicles will be greater. The residual internal combustion engine of the hybrid system runs on synthetic fuels. This is why Porsche and Audi, and perhaps also BMW, are showing an interest in (re)entering the pinnacle of motorsport.
On a slightly smaller scale, Smedley is starting its own low-cost racing series in 2020: Total Karting uses electric go-karts. The series is especially aimed at beginners who have little money.