How data helped keep Red Bull’s F1 team on track during the pandemic

September 19, 2020, 10:00 AM UTC

The reception area at the Aston Martin Red Bull Formula 1 Racing team’s headquarters in the English city of Milton Keynes, a commuter-belt town about 55 miles north of London, features a soaring two-story display case packed with silver, gold, and glittering glass cups, tombstones, and trophies—a visual, in-your-face reminder of the team’s prowess since the energy-drinks company took over the team in late 2004.

Last season, the team’s top driver, the brash 22-year-old Dutchman Max Verstappen, finished third overall, behind Mercedes’s two drivers, Valtteri Bottas and Lewis Hamilton—who, for most of the past decade, has dominated his sport the way Michael Jordan or Tiger Woods once dominated theirs.

The Red Bull team thought they were coming into this season with a shot at displacing Mercedes to win the championship for the first time since 2013. But that was before the coronavirus pandemic struck, and turned this season, as it has for every professional sports team in the world, into the strangest season in living memory.

The 2020 F1 season was about to kick off with the Australian Grand Prix (GP) in Melbourne in mid-March when the race was suddenly canceled after a member of the McLaren racing team contracted the coronavirus. Then, as the pandemic gathered pace, the season was suspended for four months.

As was the case for most professional sports, the long pause took its toll. Formula 1 itself—which American billionaire John Malone’s Liberty Media purchased in 2016 for $4.4 billion—reported that its second quarter revenues plunged to just $24 million—compared with $620 million in the same quarter in 2019. That contributed to a $122 million operating loss during the quarter. Liberty injected an additional $1.4 billion in cash into the business through an asset-swap, and it amended the terms of a $2.9 billion loan and a $500 million revolving credit line to keep the sport operating.

Weathering the worst

Individual teams suffered too. Last year, in the second quarter, F1 shared some $335 million in profits with the racing franchises. This year, they got zilch. Red Bull—which made $1.1 million in operating profit on revenues of $314 million in 2018, the last year for which financial results are publicly available—is well capitalized by its owner, the energy drinks company Red Bull, and weathered the pause better than some others. McLaren was forced to raise new capital. One team, Williams Formula 1 Racing, family-owned for 43 years, was forced to sell out to U.S. private equity firm Dorilton Capital.

When racing resumed again with the Austrian Grand Prix on July 5 there were plenty of changes to contend with: There are no fans, sponsors, or guests at the tracks, and only a few authorized members of the media. Teams are allowed to travel with a maximum of 80 staff, with the number allowed on the track prior to a race—when drivers, coaches, engineers, and mechanics from the various teams are often packed close together—restricted to half that number, dropping to just 16 people in the final three minutes before the start. Masks have to be worn at all times at the track, and each team member working in the paddock or pit lane needs a negative COVID-19 test and is retested every five days that he or she is on the road. Teams are required to remain in strict social bubbles. When not at the track, they have to be sequestered in hotels.

There are plenty of changes back at Red Bull’s home base in Milton Keynes too. On the day I visited its eight-building campus, masked employees followed one-way systems, carefully delineated with yellow-and-black tape on the floor, past hand sanitizer stations and signs reminding employees to maintain social distancing. Staff canteens have been restricted to takeaway items only. To minimize any risk of spreading the coronavirus, the workshops where engineers assemble the team’s race cars were off-limits to outsiders. Instead, I was ushered into the team’s cavernous, museum-like events space in a building known as “MK-7.” Here, spotlit examples of every car Red Bull has raced since 2005 are displayed in a fanlike arrangement. Even motionless, their engines silent, they project an energy that belies the room’s hushed stillness –like a pack of lions poised to pounce.

Here, in a glass-walled conference room, Matt Cadieux, the affable, sandy-haired American who serves as Red Bull’s chief information officer, tells me that one of the biggest challenges that Red Bull and other F1 teams faced with the coronavirus is far less visible than absent fans and ubiquitous masks.

220 mph

A melding of a man and machine, F1, more than almost any professional sport, depends on data. As the drivers scream around the track at speeds that can exceed 220 miles per hour, sensors monitor every part of the car, streaming performance figures back to the paddocks, where engineers scrutinize it in real time, looking for telltale signs that motors are overheating, parts are wearing down, or tires are losing their grip. That analysis can make the difference between winning and not finishing at all.

All this data is further scrutinized between races, as engineers puzzle over ways to eke out precious more milliseconds of speed from their multimillion dollar chariots. “We introduce new components targeted for specific races, and we can execute lots of engineering changes throughout the course of the season,” Cadieux says.

But this year, the long pause in racing owing to the coronavirus meant the teams didn’t have that data. “By July, we would normally have accumulated hundreds of laps worth of data from race weekends to help develop the car. But without this physical running, we had to rely heavily on simulated data,” Christian Horner, Red Bull Racing’s team principal, the equivalent of the chief executive officer, says.

One of the prime ways F1 teams generate simulated data is though wind tunnel testing with 40% to 60% scale models. The teams often tweak the car to match the track layout, with the number of straights and corners, and the tightness of those bends, as well as the type of track surface and expected temperatures affecting how a team will tune a car’s suspension, manage its particular tires, and set up the car’s front and rear wing configurations. Those adjustments are tested in the wind tunnel prior to racing.

But to keep racing competitive among teams of varying financial resources, F1 restricts the number of wind tunnel tests a team is allowed to conduct to no more than 65 per week. This year, each of those wind tunnel tests mattered more than ever, Cadieux says. “The importance on the analysis, the insight we get, was more this year because we went a longer period of time without having real physical test data from the race track,” he says.

Running the wind tunnel was more difficult too. The team’s aerodynamics experts would normally have been on the site of the wind tunnel, which is in the English town of Bedford, during the test. Now, owing to COVID, the wind tunnel operators had to be in socially distant bubbles from the analysts, who monitored the tests remotely from Milton Keynes, Cadieux says.

Red Bull worked with data analytics firm Kx, a subsidiary of First Derivatives PLC, to maximize the value it could get out of its limited tunnel tests. Kx is best known for its work in finance, where it helps the trading floors of banks and hedge funds analyze real-time market data. “We hadn’t worked with a Formula 1 team before,” says James Corcoran, the chief technology officer for financial services at Kx. But he said the demands of the racing team and a fast-paced trading floor were surprisingly similar.

Before Red Bull started working with Kx in 2017, Cadieux says, data from a series of wind tunnel tests was gathered, uploaded to a high-performance computing cluster, and analyzed in batches. Kx allows engineers to view the data and analytics in real time, while the test is underway, and make adjustments between each test. The platform also lets the team view and analyze historical data it has stored.

Trouble in Austria

But even with such sophisticated systems, wind tunnel testing and computer modeling can only do so much. There remains a disconnect between these simulations and the real world that is not easily bridged. And when Red Bull finally got a chance to put its two 2020 cars on the track at the first race of the COVID-abridged season—the Austrian Grand Prix held in Spielberg—it was in for a nasty surprise. Hamilton’s Mercedes collided with the Red Bull driven by Alex Albon, spinning it around and taking it out of the race with less than four laps to go. Meanwhile, Verstappen’s car suffered a power failure and did not finish the race—a preview of problems with Red Bull’s Honda-built engines that would continue to plague the team throughout this year’s season.

Ultimately, Cadieux says, there’s no substitute for real races. “In general, early on in the season you have a lot of uncertainty, and you learn a hell of lot in the first few races,” he says. “The first race wasn’t what we expected, but we recovered relatively quickly.”

He says that the key then is to take data from the track and feed that back into the engineering process. “If you’ve done a good job designing the car as a platform,” he says, “it is a few tweaks and you are back on track.”

That statement could just as well be a metaphor for Red Bull this year. After the abysmal first race, it quickly recovered. Its two cars finished in third and fourth in the second race of the year, the Steiermark GP, held a week later on the same track. Its top driver, Verstappen, came in second the week after that at the Hungarian GP.

In a normal season, most F1 races are held two weeks apart, giving teams more time to make adjustments between each. This season, F1’s compressed racing schedule, designed to make up for time lost during the pandemic, means the teams had to compete in an unprecedented six races in just seven weeks—and had three back-to-back races between the end of August and mid-September. “It means that once you get the real race data, get your head around it, update your models and simulations, the amount of lead time you have to actually design and manufacture new parts is compressed, so that adds to the challenge of developing the car,” Cadieux says.

Setting up the car for each race involves a complex web of decisions. “There are hundreds of variables, mechanical but also with software, that you can use to adjust the configuration of the car,” he says. “And understanding all those permutations, earlier in the season, there’s more uncertainty. As you get more experience running the car, it becomes an easier task.”

Going into the Russian Grand Prix in Sochi on Sept. 27, Verstappen is now in third place among F1 drivers, having won the F1 70th Anniversary Grand Prix and managed six podium appearances.

The team is in second place overall, with eight races left in the season, which ends with the Abu Dhabi Grand Prix on Dec. 13. But given that Red Bull trails the superpower Mercedes team by more than 150 points, it seems likely that its dream of recapturing the championship will have to wait another year.

More must-read international coverage from Fortune: