Mercedes-AMG Petronas Motorsport: Setting Up for Success by Expecting the Unexpected
Confidence, clarity, communication, and collaboration
It's no secret that data is crucial in Formula One™. Six-time FIA Formula One™ World Champions Mercedes-AMG Petronas Motorsport is known for continuous evolution and innovation—and its data is especially critical when it comes to car setup. Teams sift through billions of parameter combinations to find the Holy Grail: the fastest possible setup combination for that track, that day, that car, that driver. And with so many variables, it would be impossible to be fully prepared for what you don't know. But here's where Mercedes-AMG Petronas Motorsport shines—in expecting the unexpected. That ability, to find the winning combination, doesn’t come without trial and error.
Identifying optimal configurations in highly multivariate data requires sophisticated visual analytics and data science analytics tooling, especially because, as the race weekend unfolds, the team of strategists and engineers have limited time for iterative cycles. In F1, conditions can change faster than the blink of an eye, so while the team can determine most settings, it’s the combination of external events that present challenges and opportunities.
Throughout the season, participating F1 teams travel to 21 circuits around the world. Each circuit has its own unique set of characteristics; from narrow street circuits, such as Singapore or Monaco, to open tracks, like Monza.
The team responsible for setup is the performance group, led by Andrew Shovlin, trackside engineering director. From the collected data, it’s up to Andrew and his team, including Senior Performance and Simulation Engineer Michael Sansoni, to go through all the possible car configuration combinations to make sure they extract the best possible performance from the car over the race weekend. The group also needs to ensure that when the drivers provide their feedback, the team can be responsive and apply the information received to car setup by the time qualifying arrives on Saturday. Setup has big payoffs: for example, getting to the front row of the starting grid. The higher a driver starts on the starting grid, the better the chance they can challenge for the lead throughout the race.
Setup for Success
Imagine finding a needle in a haystack. That's what it's like trying to find the right setup, the one that will make for the fastest car, from the billions of possible car configuration combinations. As tires wear and the track conditions change, teams can tune suspension, aerodynamics, and power unit settings to affect performance. Potential lap times, tire degradation, and tire life grip and wear are also part of race strategy.
"We need to make sure that we get the perfect combination of each one of these different identities, and that's really what we leverage TIBCO for, and where we really see the performance benefits," said Sansoni.
To prepare for a race weekend, the team runs millions of simulations, each one trying to predict various situations during the race weekend. These simulations take into account environmental factors such as changes in weather and track, including the many changes to the circuits year-on-year. In the world of F1, nothing is constant, so it's crucial for teams to constantly monitor these factors.
"Getting the car setup right is quite difficult because it's never a precise activity. You’re dealing with a lot of unknowns. You don't know how the track will evolve during the weekend, you don't know what the weather will be like on Saturday and Sunday, and you're also having to broadly get the setup correct for both qualifying and the race, which are two very different regimes," said Shovlin.
Not only does the team try to predict the fastest car setup, it has to provide the strategy team, the mechanics, and the drivers, with a set of alternative changes they can make based on the feedback received. This feedback loop gives Mercedes-AMG Petronas Motorsport the opportunity to continuously improve during every lap on the race track, whether it be during Practice, Qualifying, or the Race itself.
There's No “I” In Team
Like in all other areas of F1, collaboration is key to success. The team is limited in the number of personnel who can operate at the track, which means they need to collaborate between the track and the factory. Collaboration also needs to take place within various sub-teams who are all specialists in their respective areas. It’s up to the team to extract everyone’s expertise, at just the right time, to create maximum impact on the car’s performance.
Teamwork starts with the simulator group and a proposed setup. Once the setup is run through the simulator, it is evaluated with input from engineers and test drivers. Performance of the car configuration is gauged, and improvements identified.
All of the work that is done prior to a race weekend acts as a rehearsal for the actual event, with all the pieces in place to maximize performance on the track. Several groups are involved, including the wind tunnel team that provides aerodynamic data as inputs to the simulations to ensure the best aerodynamic performance from the car.
If at First You Don't Succeed, Try Again
Mercedes-AMG Petronas Motorsport faces two major challenges when it comes to setup. The first is the sheer volume of data that needs to be processed. The second is overcoming the cadence of data. As the sport continues to change, teams that don't evolve quickly won't be successful. So, for Mercedes-AMG Petronas Motorsport, one of the keys to its continued success has been continuously innovating and evolving alongside F1.
"We need to be able to react with all the new data we receive, the new aerodynamic platforms we receive from aerodynamics, the optimum performance from the suspension and power units,” said Sansoni. "We need to be able to combine these as soon as we have the latest data to ensure, that out of those billion combinations, we can choose the fastest setup."
Another challenge outside of data is the rules enforced by the FIA. One such rule is using the same setup during qualifying and race. Once the setup has been chosen for qualifying, there is no going back, what’s chosen carries the team through race day. So, the key is to optimize the setup performance prior to the qualifying session, using data gathered from the simulator and historical data from the previous race or circuit from years past.
"Each team is trying to get the absolute maximum amount out of each race weekend, and we have a limited number of tires, a limited amount of time, and a limited number of runs that we can do," said Shovlin.
Three C's: Confidence, Clarity, Communication (and Collaboration, Makes Four!)
While generating data is very easy for a team like Mercedes-AMG Petronas Motorsport, what is challenging is understanding it, which is what makes for a championship-winning team. Every improvement made needs to come from the data analyzed, with consensus amongst the group of specialists. A consistent view of data is important, as is having faith and trust in the data and the analysis produced. When looking ahead to a race weekend, Mercedes-AMG Petronas Motorsport has two uses for TIBCO Spotfire and TIBCO Data Science software. The first is analyzing previous races and previous circuits. The team can apply predictive algorithms to understand what changes were made at previous races to help them learn and predict what to do at future events. Each circuit is unique, with characteristics that the team can learn from.
"A lot of the work is just checking that what’s provided is building up to the event and that it isn’t going to mislead us in any way,” said Shovlin. “We’re constantly refining that because models are getting more complicated."
The second use case focuses on dissection of the large magnitude of simulations to extract the optimum car setup and discover performance trends that help the team react to what it sees during a race weekend. The team no longer needs to wait on the latest data in order to send out an analysis report. Once the team accesses more complete information, they can continually develop analyses, provide pre-event predictions and information. Based on these predictions, stakeholders are able to create a picture of the race and understand, with confidence, that they have the best representation of the car for a set of anticipated conditions.
"There are so many parameters you can change on the car to deal with wet, dry, hot, or cold conditions. A lot of that comes from experience, but when you can run millions of simulations, and use a tool like Spotfire software to distill the results into a subgroup, you can conceptualize and understand what’s really going on,” said Shovlin.
The adoption of Spotfire and TIBCO Data Science software has significantly changed the way members of Mercedes-AMG Petronas Motorsport consume some of their data. Data is received faster, leveraging the expertise of all extended teams.
Fail Fast, Succeed Faster
If you work for an innovative company, you know the term "fail fast and fail often." What was once taboo is now the guiding principle that drives company culture. In fact, failing fast and failing often is the best-kept secret in digital transformation, imperative to a business’ success. It creates innovation by giving the freedom to think beyond how things have always been done—and to disrupt and change for the better.
However, fail fast and often needs to be done in a highly quantitative manner, with appropriate analytics tooling and within the context of examining potential outcomes, value theses, and ROI tradeoffs. To anticipate the future, you need to be prepared analytically and for many potential scenarios, and know how to react quickly. This is TIBCO's DNA embodied in the Connected Intelligence platform and portfolio of products.
Mercedes-AMG Petronas Motorsport's approach uses data analytics to identify potential problems, find opportunities for improvement, and collaborate within sub-groups to make that improvement a reality. This doesn't just apply to Mercedes-AMG Petronas Motorsport or F1 in general—your company can make this a part of its feedback loop.
TIBCO Helps You Anticipate the Unexpected
You might be wondering how the work being done in a simulator for a high stakes F1 competition applies to your company. Well, the lesson learned is one of measure and response, often articulated as sense-and-respond, and this applies to many industry sectors and use cases. What the F1 simulator, and all of its surrounding technology demonstrate, is that If you are a technology or business leader, you can take a vast amount of data in a short period of time, visualize it, and run machine learning algorithms to derive insights and patterns that enable collaborative tech and business teams to make more informed decisions faster. This improves the odds of getting ahead in a very competitive business world.
For every asset and product, one can develop virtual replicas in software, of the same functionality. That is the concept of digital twins, and one reason why they are now consistently used to produce better results and deeper insights for operational optimization. More and more companies are adopting digital twins to survive their ever-growing competitive markets.
Digital twins help companies like Mercedes-AMG Petronas Motorsport address their biggest data challenges. The end results include process optimization, insights into predictive and condition-based maintenance, and optimal business action on the event stream, with the potential to lead the team to its sixth Constructors' Championship.
Click here to read the full success story: Mercedes-AMG Petronas Motorsport: Setting Up for Success by Expecting the Unexpected