Perfecting a Sport of Precision: Formula 1’s Integration of Big Data Analytics
Formula 1 is one of the most prestigious motor sports in the world testing both the athletic capabilities of drivers as well as the engineering skill of the teams who design and manufacture the cars. With advancements in mechanical engineering, the sport continues to become more and more competitive, pushing teams to find new advancements through data analysis. With data engineering teams becoming increasingly prominent in the upcoming seasons, it is clear that data analytics will continue to play an impactful role in the sport. For the decades to come, teams who can successfully leverage data to attain more points and shorter lap times will be able to push the bounds of the most technologically advanced sports vehicles the world has ever seen.
History of Formula 1
Formula 1 is one of the most challenging and arduous sports due to the sheer nature of due to the immense demands placed upon drivers to race for several hours at a time and for engineers to design vehicles that can handle the strain. With the seasons’ races taking place in twenty-one different countries across four continents every week, it requires constant adjustment and adaptation in order to succeed. It is a true test of mind and body as mechanical engineers and drivers work side by side to push their cars to new peaks of performance and excellence.
Though motorsports have existed much longer than Formula 1 has existed, the league has its modern foundations in 1946 when the Fédération Internationale de l’Automobile (FIA) was founded and made the first standardized rules for grand prix racing (Formula 1). The first World Championship was in 1950, and since then, the sport has grown from a European circuit to one of the most prestigious automotive racing competitions to date with drivers and teams participating from all over the world (Formula 1). Though this sport has maintained the majority of its presence in European and Asian countries, Formula 1 has recently gained a lot more traction and popularity in the United States. Ten teams, twenty drivers, and hundreds of engineers work together with the common goal of securing a spot at the podium at the end of every race.
There is also a more complex points system at play that rewards excellence and punishes teams for any rule violations that occur during the race. Points are awarded to the top ten drivers in every race and the higher a driver places, the more Championship points they earn for themselves and for their team. Every single point counts towards the championship, which incentivizes teams to continue to work throughout the season to decrease lap times and move to higher podium finishes with every race (RacingNews365).
Similar to the hundreds of people all working towards the same goal, the cars themselves are machines with thousands of hours of manufacturing and design behind intricate parts specifically designed to achieve that goal. In a word, this sport is meticulous. With the strict regulations imposed by the Fédération Internationale de l’Automobile, the competition becomes more difficult every year. This has forced the engineers to expand the creativity behind their technical solutions. In order to remain competitive, teams have now decided to look at fields other than mechanical engineering to maximize their points. The most successful teams have found their solution in data analytics.
The Origins of Data Science in Formula 1
Given the evolving nature of Formula 1, it’s not surprising that the sport utilizes the capabilities of emerging technologies and fields such as data science. In 2021, McLaren Racing Limited became one of the first teams to incorporate data analytics into their racing strategy (Forbes). At the time, data science as a field was still in its early stages, meaning that there was a lack of professional data scientists with experience and even fewer who also had a specialized industry knowledge of motorsports. At first, McLaren Racing began with citizen data scientists and data engineers who offered any amateur knowledge and skillsets they had to the team. Much of this analysis came in the form of visualizations and predictions for the best racing strategies for drivers to use, proving to be crucial in enabling the best decision-making on and off the racetrack. As helpful as these initial attempts were in allowing McLaren to refine their racing strategies, there was one major flaw. Without experienced data scientists who also have comprehensive domain knowledge it is impossible for these initiatives to have lasting impacts. This would lead to the rapid incorporation of full-scale data engineering teams dedicated to analysis.
The F1 Data Analysis Process
Let us dig a little deeper into how this process actually works in the background to guide the decision-making process of drivers. Formula 1 cars are equipped with hundreds of sensors that track diverse amounts of information from the conditions of the car internally, externally, and even the surrounding track environment. These vast amounts of information are compiled so that the data engineers can begin their analysis. The main reason teams utilize analytics is to gain an edge over other competitors. In a sport where fractions of a second can be the difference between victory and defeat, optimization through automated analytics has proven to be the most effective method of making those advances. Though the engineering behind formula cars is intricate, there are three key aspects that can be most quickly analyzed and adjusted during a race. These generalized areas are tire performance, fuel consumption, and aerodynamics. By detecting the rate at which the tires are wearing out in a particular race teams can preemptively communicate to the drivers when they should make a pit stop tire change rather than waiting too long and risking tire failure during the race. Understanding fuel consumption and efficiency allows teams to understand how changes to the engine affect the ability of the car to travel faster speeds for longer distances. Aerodynamics is essential when it comes to generating more downforce on the car. This downforce is critical in allowing the car to effectively grip the road without becoming too heavy which in turn allows drivers enhanced control while making sharp turns at high speeds. In some cases teams have even utilized data analytics to better understand fan engagement which in turn can allow them to generate more support from the audiences and through sponsorship for the drivers. AWS in particular has been utilized to display live car positioning which allows audiences to have a deeper understanding and experience during the race from their outside view (Medium). In some cases the technology has also been utilized to create platforms on which fans can simulate racing against other professional drivers in close-to-real-life conditions.
During the season, teams only have a week in between each race to make adjustments to their cars and strategies. This strict time constraint requires that teams make only the most crucial and beneficial changes to the car in order to gain as many points as possible. For teams to make these decisions, the data engineering teams must also optimize the data they work with. The sensors measure every detail of the car for every millisecond of the trials. The kinds of data harvested by these sensors is comprehensive, from the internal workings of the engine to the external road conditions, everything is accounted for. In a single run, data engineers can gather as little as a few hundred gigabytes of information to as large as one and a half terabytes of information. In the past, it would have been up to the citizen data scientists to clean, analyze, and visualize the data in a manner that would allow the mechanical engineers and drivers to make the most informed decisions possible. As data engineering teams have become more robust and the tool sets available to them have expanded, some teams now utilize AI-powered models. These models are designed to test out all possible modifications that could be made to the car, taking into account the car, the driver, and the next race track. Teams are then directly informed by these algorithms of the optimal strategies to use for their upcoming races. These tools have become so sophisticated to the point that certain analytics and visualizations are displayed during live streams of the races so that audiences can see what informs the critical game-changing decisions made by their favorite engineers and drivers in real time.
AI-Powered Models In Formula 1
As we have seen, data analysis at the base levels is most helpful in decision making during a race. When it comes to planning for future races, AI-powered models have become significantly more popularized for their ability to make decisions that are smarter, faster, and more efficient. There are four cornerstones in formula 1: fuel efficiency, aerodynamic downforce, tire wear, and driver training. With the help of artificial intelligence simulations all of these can be enhanced in a fraction of the time (Medium). One of the biggest benefits of AI-generated models is that they can crunch through thousands of simulations of different race scenarios and strategies before a driver even gets in the car.
The most common example that is seen is the pit stop. It is essential that drivers not only make pit stops at the appropriate times but that the stops themselves do not take longer than necessary. Every second that the driver spends in a pit stop is more time for other drivers to take the advantage so a short pit stop is crucial. With the help of AI teams have been able to devise strategies that take the absolute minimum amount of time to safely do a tire exchange and fuel replenishment so that drivers can go back just as quickly as they came.
AI is also particularly helpful in devising strategies to conquer the unpredictable.
We can consider the example of a change in strategy due to unpredictable weather conditions. Rain is one of the most difficult obstacles to overcome in Formula 1 particularly for the way it affects how tires grip onto the track, the behavior and visibility of other drivers, pit stop strategy, and the conditions of the entire track. Prior to these models, if it began to rain during a race the driver would have no choice but to pit stop as soon as possible to more durable rain tires and much of the pre-planned racing strategy would have to be scrapped. From that point on it would be left to the driver to race based on how they experience changes because of the rain. With data-driven AI simulations teams can pre-plan alternative strategies.
In the week that passes between each race teams are able to use virtual modeling and other simulations to pinpoint the weakest parts of their performance (Medium). This knowledge saves not only time, effort, and financial resources but can also be life-saving as it can help engineering teams detect potential component failures ahead of time. In these scenarios teams are able to ensure that every car starts the race at its peak form while also understanding how that performance changes during and after the race.
Outside of the racing world these models are also helpful in enhancing the Formula fan experience. One of the biggest proponents of this has been AWS which has recently begun a long-term partnership with Formula 1 (Amazon). Utilizing live feeds from 300 sensors which produce 1.1 million data points per second, teams are now able to see the displays of analytics that drivers and teams use when coordinating heart-stopping real time decisions. Fans can now have a deeper understanding of what goes on in the background of a race and follow the footage of their favorite drivers in real time. AWS has also created something called a Battle Forecast which has been particularly exciting for viewers. One of the key opportunities a driver has to overtake another car is during turns as it is often too difficult to gain the necessary speed during straight-line stretches of a track. AWS software has been utilized to predict in real time whether or not a driver will be able to successfully get past the car in front of them by using analysis of previous lap times and the previous paths taken by the driver on that same turn. This displays the power of data analytics not only in prediction but in quick and effective real time analysis.
Moving to a More Sustainable Future
A core value of Formula 1, motorsports, and engineering is constant evolution. These knowledge gained from these new data driven strategies now allow teams to make decisions about future versions of the car for future seasons. As the industry begins the transition to electric vehicles, data analytics will begin to play an even more important role. Though it is less popular due to its recency, there now exists a Formula Electric circuit, better known as Formula E. While many teams and engineers prefer the traditional foundations behind Formula 1, Formula E signifies a future that tests new technologies and challenges the industry to work towards zero emission vehicles (EnelX). Formula E was created in partnership with the United Nations Environment Program to highlight the power and benefits of electric mobility (EnelX). To follow with the industries, FSAE has also begun to follow suit with the emergence of many Formula SAE Electric teams across hundreds of universities. After more than fifty years of internal combustion engines, teams will now have to find a way to maintain the advantages they already have with a completely different kind of car. Data will be essential in allowing them to rethink the wheel without reinventing it.
