23 races, 21 countries around the world, and speeds up 248 mph (400 kph) — while Formula 1 is a sport like no other, many of the racing teams face the same challenges as other businesses, especially when it comes to data and analytics. Take a behind-the-scenes look at McLaren Racing’s formula for analytics and R&D success and get tips for tuning up your analytics.
What does the analytics process at McLaren Racing look like?
The data science department will provide tools and advanced analytics capabilities to other teams so domain experts can perform their own analytics. The technical team and race engineers look at driver data, vehicle dynamicists analyze suspension and the tires, and strategists dig into game theory and how we’re interacting with other teams. It’s really powerful when domain experts have self-service environments where they can access and explore the data. It helps them find answers simply and quickly and frees up data scientists to think about the next cutting-edge tool we can develop to magnify impact.
How does McLaren Racing approach modeling?
We build models of the car that undergo stress testing and structural testing to make sure the car can withstand temperatures and forces. Simulations include computational fluid dynamics, aero, and laps of all the tracks. All of this is virtual though. When the car is tested on the track, we might see big deviations from what we’re expecting.
We’ve found that the models that are simpler are oftentimes a lot more powerful than complicated models. Simpler models help us get to better solutions since we can more clearly understand why a model is reacting in a certain way.
We improve our models by connecting with the domain experts who will be using the models or interpreting the results to discover what tools they need.
What lessons learned from analytics do you have?
Although we’re in F1, we share many of the same challenges that all businesses do like maintaining a data culture.
We’ve found that for data science to work across the whole company, everyone needs to understand and appreciate the importance and potential impact of what data science, analytics, and machine learning can do.
As an R&D industry, we also need to balance short-term projects with long-term R&D. Many of the initiatives that we try aren’t going to work or if they do work, they might not bring enough benefit to us to be worth supporting. We do many short-term experimental projects before we commit to longer projects, which might involve writing quick scripts, testing the data, and attempting to reproduce results. We don’t have a large data science team so we’re careful about how we spend this resource.
What areas outside of racing does McLaren use data?
Our CEO Zak Brown often says that McLaren Racing is built on three pillars — our fans, partners, and people. We use analytics in all three of these pillars.
Fan engagement is critical because without fans Formula 1 wouldn’t exist. We always want to make sure our fans feel engaged and can access the team, our social media, and events. We also strive to work well with our partners and ensure that engagements benefit both sides.
Within our own team, we want employees to feel valued. We use analytics to determine the mood within the team. Understanding our own team is just as important as understanding the race car. We wouldn’t have a fast race car without a team that works well together.
Over the last few years, we’ve seen the positive impact that looking after these three pillars has had on race performance. Even if they’re one step removed from making the car faster, they have a big effect on the whole journey of the team.