Motor racing is a fascinating sport with the most balanced of foundations. For many sports, cars are standardized across every racing team—all built with the same major parts and configurations. A fraction of a second can be the difference between winning and losing. This means driver ability, strategy, and car configuration are key —and even small incremental improvements can greatly affect the outcome.
I have been traveling around the USA as part of DataRobot’s sponsorship of the RKM Racing team, which started last year as an Andretti Autosport Indy Lights car, and continues in 2022 with Lamborghini in the IMSA races. Through immersing myself in the sport, chatting with engineers, and exploring what it takes to win with driver Robert Megennis, I learned all about the impact of technology, people, and data in modern racing.
The Race for Better Data
Author Ari Kaplan on the track at an IMSA race.
Data has existed in motor racing for a while, but the volume of data has been accelerating at an unprecedented rate—as is usage of AI and machine learning. This brings about new approaches and opportunities for gaining an edge. For example, racing simulators have become a major source of data, from the units drivers have in their private homes to more advanced models operated by large and established manufacturers such as Ferrari and Renault.
Whichever simulator a driver uses, the principle remains the same: letting them drive virtually in any car, with any configuration, and on any track. This includes user-defined racing conditions, right down to how many cars have driven on and worn down the virtual surface. All this helps drivers better prepare for—and react – during live races.
Robert told me that simulators are very effective to help him prepare in some regards, but they lack “physicality”—there’s no sense of movement through the body, no g-forces of rapid turns or acceleration, and no real feelings of danger. This can also bring advantages: “I can do a 20-lap run and go faster and faster until I destroy the (virtual) car and then reset the simulator. It lets me find those limits.”Simulated data is pored over by analysts to provide actionable advice for upcoming races. With real-life track time being limited, but conditions being predictable, it’s useful to know in advance how best to approach a drive. Robert explained that preparation isn’t solely about raw data, a car, and a race weekend—it involves the long-term and digging deep into human behavior: “A big part of racing is being smart for an entire calendar. And because you’re racing the same people for months, you need to know how to get the most out of them—who’ll let you pass and who’ll run you off the track.”
The Need for Speed
A photograph of the IMSA Lamborghini car that Robert Megennis drives.
Technology continues to advance, unlocking new insights, new use cases, and new strategies. New data needs both human intelligence and artificial intelligence to find signals in the noise— determining patterns and nuances of the information in order to provide new competitive advantages. Modern Indy Light and IMSA cars have over 300 sensors that collect data in real-time, thousands of times per second throughout the entire race. Simulators churn out data at the same relentless pace.
“There’s so much from an AI side I don’t see, I think because engineers don’t want me to start overthinking how I drive,” said Robert, adding that team priorities are logically applied: “The driver’s job is to drive the car to the best of their ability, and the engineers engineer the car.”
Drivers, therefore, concentrate on a subset of data to help them with race strategy and car handling, while engineers delve into insights on improving car performance, such as increasing downforce and reducing air drag. All aspects of racing can be explored through data, including comparative lap times with other drivers, tire RPM, speed, gearing, steering wheel angles, suspension levels, and braking.
“Brakes are particularly important,” explained Robert. “Because everyone’s in the same car, with the same engine and tires, winning is down to how you drive the car and how the engineers set things up. When we look through data, we first examine the brake trace, because racing is gained and lost on the brakes.”
During a race, automating and streamlining this process is essential to process and react in real-time. ar, weather, track, and competitive conditions can change without warning, and insights are needed as quickly as possible so teams and drivers can adapt while there is still the opportunity. Being able to correctly interpret data at speed is vital, but there’s too much for humans to process in real-time. Here, AI can aid teams, helping them identify and adjust tactics to account for wind, track conditions, air temperature, fluid levels, tire degradation, air intake, and more.
Faster Decisions with Data and AI
The Lamborghini in action on the track.
Making data digestible is crucial in any industry—people without PhDs also need to be able to gain valuable insights. In car racing, the driver’s brain is already making countless decisions every second—and at high speeds. So, what begins as myriad individual pieces of data must be intelligently filtered down to a small number of levers that can be pulled or into important information a trackside team can quickly relay to a driver over the radio. Success stems from agility, the ability to make decisions on the fly, and fusing technology with an understanding of people, being able to consider the performance of other drivers as well as the condition of your own.
Win or lose, data continues to be invaluable post-race. Failure is an opportunity to learn how to do better over the long term. Success demands that analytics understand why your team’s driver outperformed the others in order to repeat that success. Opportunities come from exploring new ideas and setting the right expectations. Debriefs include elite groups of people spending time pouring through data, video, and constructive and collaborative conversations among the drivers, pit crew, engineers, and strategists.Since data can pinpoint when a driver is inconsistent or tired, this incorporates human elements—– even if they don’t know it.
“This is important,” said Robert. “Fade at the end of the race—lose concentration for one second or make one tiny mistake—and you’ll fly into a wall at 200mph.”
The Human Advantage
When I spoke to Robert he was unsurprisingly bullish about the role of data in Indy Lights, IMSA, and all car racing circuits. “A team’s data is its key to success,” he affirmed. “It’s how we win races. Without it, there’d be no way for us to do well, because we wouldn’t be able to pull insights from that data and know what has to happen for us to win.”
What I found interesting—and something that’s often lacking in the discourse around data—is how much the human angle pervades everything.
AI is more than cold numbers. It informs decision-making, but requires human analytics and context to help you gain a competitive advantage. This works best when humans and AI cooperate and work together in step. AI excels at doing repetitive tasks and can find deeper nuances and complexities in vast amounts of data. Humans decide which questions need to be asked and guide how to iteratively adjust the insights with real-life applicability.
Robert also told me he thinks looking deeper into people—modeling human behavior—is the next step for AI in his sport. “We have everything we need to know about the car, so data will become more about the driver,” he said. “You’ll analyze my body and what’s going on in my head. When looking at my performance, you’ll know my heart rate and even what I’m looking at.”
Along with providing further advantages to racers, this rapid evolution of data usage and AI will impact team composition. Today, Robert’s team includes an assistant strategist who trawls through data traces to confirm that everything is running as optimally as possible. “We’ll see AI take over those processes and make them simpler, so people can focus on people things, like driving, engineering, and making physical changes to the car,” he said.
Opportunity Through Collaboration
AI is still early in the racing industry, but it’s already showing great value. Alongside more plentiful and diverse data sets, it’s helping make the car as fast as it can be—and the driver as good as they can be. Success comes through augmented intelligence – where human intelligence and machine intelligence collaborate.
This union is key, not only in sports, but in every industry. In today’s era of informational overload, success comes from a combination of obtaining data, people who understand the business and the data, and leveraging AI to gain insights to illuminate what’s missing and ultimately how to be best positioned to win.