Data engine: McLaren's Lando Norris and Oscar Piastri on their F1 data and AI edge

In Formula 1, data isn’t just the new oil — it’s probably more important than any other component save the driver. Every megabyte of data that F1 racing teams collect could hold the key to a points-winning finish or a spot on the podium.

At McLaren Racing, the team is not only hard at work every second of every day to achieve a perfectly optimized car, they’re also fast tracking data and AI innovation (as we wrote about this summer) to maximize their race performance.

As McLaren’s racing stars, Lando Norris and Oscar Pisastri, gear up for the final race of the season, and a good shot at securing the Constructors’ Championship, we sat down with the two drivers to discuss how cloud and AI technologies have helped the team dig into its data.

It’s this AI-powered analysis, which can reach hundreds of millions of simulations before each race, that has helped deliver winning strategies and record-breaking pit stops, improved driver performance, and informed development of the fastest cars possible. It’s all in the service of every marginal gain and competitive edge possible to put McLaren challenging for a world championship.

Let's kick things off by discussing how you and your teams leverage data during a typical race week.

Lando: Sure. It’s a lot of data, and we tackle it in a few stages: pre-weekend, arrival into the weekend, and then the weekend itself. With all that, we have to simplify it as much as possible to focus on the most critical parts. Between my personal team and myself, we analyze data from past weekends and even past years to prepare. This includes reviewing videos and data from previous races to understand my strengths and weaknesses, and using the simulator to compare my performance with Oscar's.

We also use data about what the track will be like, the weather forecast, the tires we have, and more to generate predictions for tire strategies, lap times, and more. All of this helps us create the best possible plans for practice, qualifying, and the race.

Oscar: Yeah, so it really depends on the time of the season and the type of race, but preparations usually begin on Wednesday or Thursday, where we review old videos, analyze past performance at the circuit, and start looking at setup options. On Thursday, we have media commitments and engineering meetings to discuss setups and run plans for practice sessions. Friday is practice day, where we experiment with car setups and gather as much data as possible for qualifying and the race.

McLaren Racing runs close to 300 million race simulations prior to a race. Can you walk through some of the specifics of how your teams use simulations to predict the most effective race strategies?

Lando: The team runs simulations before the race to understand as many variables as possible. We want to give ourselves the best opportunity to be prepared. If anything happens in the race, we can lean on the information we already have on what direction to take with the tires or if it’s going to rain. It also helps me to know how we think our race is going to pan out before even getting in the car, so I can adapt and drive accordingly throughout the weekend.

These predictions are constantly being updated, too: after practices, qualifying results, and through into Sunday. Whether it’s weather, temperatures, tires, how I’m feeling in the car, how we set up the car — all these bits of information go into making predictions that help us set our plan of attack.

And how do simulations help inform how your cars are set up?

Oscar: You know, obviously, we can't drive the cars 24/7, so simulations help us generate a starting setup. Of course, there is some human finesse in there — we don’t always start with exactly what the simulations predict for us, but it’s always a good starting point.

This data is also channeling into the upgrades we put into the car. That’s been a real strength of ours in the last couple of seasons, bringing upgrades that work very well and have a really good correlation from the wind tunnel to the track. That’s what has ultimately got us into the position of fighting for the Constructor’s Championship.

Lando: Yeah, we basically run an exact simulation of what we’re expecting at the track, inputting different values and getting feedback on them. Even more important, but harder to do, are simulations for a track we’ve never been to before. We have simulations that run, say, car speeds and give us a lap of data to tell us the performance between two different wing levels, including track conditions like temperatures and grip levels. Running simulations allow us to preempt a lot.

So, for example, the Austin Grand Prix is an extremely bumpy circuit. From a setup point of view, you need quite a different approach. Making sure I'm comfortable and not getting rattled around too much is important. Normally, you want the car to be lower and stiffer, but you have to make compromises because the Circuit of the Americas is built on man-made land. The track changes every year. So, we look at past years again to see how things have evolved and data from other racing categories that go there like the IndyCar series. We use this data to change and alter our setups before the weekend even starts.

Tire strategy is a crucial part of the race, both leading up to and during it. What role does data play incoming up with an optimal strategy?

Oscar: Yeah, tire strategy is a massively important part of Formula 1 racing. One of the biggest challenges is that we often have to predict what tires we want before the weekend even starts. We have three different compounds of tires — from hard to soft — so we have to choose which ones we’ll bring forward to the actual race on Sunday. You can only choose a certain amount of each compound, so picking the right ones is critical. We have a ton of data around the tires and how we think they will perform on a given track, most of which comes from pre-event simulations but also based on what we see during practice sessions, qualifying, and the race itself.

Lando: Over the course of the weekend, we take in a lot of information, photos, and compare our tires on a lap-by-lap basis. During a race, this needs to be done within a matter of seconds, not minutes — that’s already too slow. So, finding a quick and easy process to look at the tires, compare and review them leads back to my engineers and to me, so we can make split-second decisions. This is one of the biggest areas with tires and strategy that has helped us develop over the last three years.

How do you think this season’s data will help optimize the car and maintain the same competitive edge next year?

Oscar: Since we’re towards the front of the standings this season, the rules say that we get a bit less time in the wind tunnel, giving the other F1 race teams an advantage in how they rethink their cars and strategies for the 2025 season. This means it will be crucial for us to get the most out of our data during pre-season testing. We get three days to get it right.

That’s where the data from 2024 will really drive the car that we develop for next year. We’ve already started working on next year’s car halfway through this season — all of that development is pretty much driven by the data and the numbers we’re seeing on the track and what’s working there. And then, in the off-season, we’ll layer on what we collect in the wind tunnel and from our simulation tools.

McLaren set a new world record for the fastest pit stop during the 2023 Qatar Grand Prix. What role did data and technology play?

Lando: Our 1.8-second pit stop is the quickest ever recorded in a time when there’s a lot of measures in place to actually slow pit stops down. Tires are heavier, wheel guns are slower and less powerful — yet pit stops are quicker than ever before. This has been achieved with a lot more data, reviews of pit stops, and practice. At the track, the pit crew can review pit stops, going through every single part of a pit stop. There are three guys per wheel and they can look at each one individually and compare them to one another. They are also able to overlay videos of previous pit stops to see what they’re doing differently.

How do you see data continuing to shape the future of Formula One, both on and off the track?

Oscar: I think data will continue to shape Formula 1 massively. For the drivers, it’s a critical tool. We can now see pretty much every driver on the grid, so it helps drivers trying to improve as well as when the team is trying to recognize weaknesses of the car. That could be high-speed corners, low-speed corners, or being slow on the straights.

Instead of trolling through a bunch of numbers and trying to guess, we have all the data and evidence there. So, data will always shape our decisions, our development, the upgrades we make, and ultimately, the performance. Data’s a critical part of our job.