How Formula 1’s McLaren team is using A.I. to fuel performance

Drivers go into each competition with a well-laid strategy, determined by their experience, past performance, car capabilities, and challengers. The drivers stay in constant communication with their team managers and engineers during any given race. These team members, sitting at a control station, monitor everything from the condition of the tires to upcoming weather as well as provide drivers with a bird’s eye view of their race positioning. They also strategize when the car is in a takeover position to move up on the pole.

Those same teammates sit down for a postmortem after every race to go over what was done well and what could be improved, and, yes, to pore through an immense amount of data collected during the competition.

“There is a lot of technology around Formula 1, more than in any other category of racing or any other type of car,” explains Lando Norris, a Formula 1 driver who competes for McLaren. “The speed and the adrenaline is something which you basically never feel in anything else but a Formula 1 car, and there are key things which I need to see—so much vital information—to perform at this level.”

Norris is referring to the feedback loop among driver, car, and the team’s mission control that makes, or breaks, every race. Norris explains that his team captures data on his vehicle, but also that of his teammate, Daniel Ricciardo, and those of the other racers from competing teams such as Ferrari or Red Bull. While Norris is relying on intuition and the “feel” of his vehicle, he is also listening to these stats and recommendations to make the most informed decision at any second during a race.

For Norris “every bit of data is changing the outcome of the race,” he says, constantly adapting while behind the wheel.

Racing at speeds of more than 200 miles per hour requires extremely quick decision-making, which is why McLaren uses Splunk, a software platform designed to analyze big data. James Hodge, Splunk’s chief strategy adviser, explains that Splunk streams data in real time from each race car’s 300 sensors, which monitor details including fuel levels, tire pressure, speed, and battery health. Those sensors generate more than a terabyte and a half of data each race weekend.

Over the course of a Formula 1 season, it tops out at 11.8 billion telemetry data points, derived from more than 440 car configurations and approximately 6,300 track laps.

That data travels in real time to live dashboards at two important sites: McLaren’s trackside “IT rig” used by on-site team members at tracks such as Melbourne, Abu Dhabi, and Monza, Italy; and to 32 engineers at McLaren’s Technology Centre in the U.K. Hodge notes that it’s “safety-critical to receive real-time telemetry.”

“No human could hope to make sense of 1.5 terabytes of data across 300 sensors each week,” Hodge says. “McLaren’s relationship with Splunk is essential to ensuring that data gets to the right people at the right time, ready for analysis.”

Splunk uses artificial intelligence and machine learning to capture, index, and correlate data, then present it on a searchable dashboard with graphs, reports, alerts, and visualizations. It also records historical information to establish data baselines and detect future deviations. All this equips drivers, engineers, and team leads with valuable information to optimize decisions for car design, driver performance enhancement exercises, and upcoming race strategy.

Norris says that learning to understand the “squiggly lines” on a screen at a young age has positioned him to be particularly data-driven when it comes to, well, driving.

The data shows Norris acceleration patterns, velocities, tire temperatures, and other relevant metrics like whether he is braking too early or too late. It presents calculations computed from 10,000 different raw parameters, distilled in a way that Norris can utilize for future races and practices. In fact, between races, Norris spends time with e-sports simulations, also plugged into Splunk, to prepare for the rigors of different tracks and improve his techniques throughout the season. 

“These bits of data shape a driver and can make him either a very good driver, a good driver, or a bad driver—it’s just how you use it,” Norris says. “The more we can advance in every area, the better and more competitive we’re going to be.”

Hodge explains that Formula 1 has always been about maximizing every millisecond, and now quick analysis of data is making that even more of a reality for the sport. It’s supporting drivers by giving them more confidence in their game-time decision-making, increasing safety by rapidly detecting and resolving problems, and preserving the engineering integrity of the cars.

“As a driver, our job is to just go out and go quicker,” Norris says. “Looking at data can sometimes be the most simple thing, and it’s those kinds of things that make a big difference to us.”