Full Transparency
Our editorial transparency tool uses blockchain technology to permanently log all changes made to official releases after publication. However, this post is not an official release and therefore not tracked. Visit our learn more for more information.
While the business world has been swooning over “big data” for less than a decade, the importance of data in racing isn’t anything new. Originally, the tools of the trade were a clipboard, stopwatch and pencil. Chris Simmons, Lead Race Engineer of the No. 9 Target Chevrolet of Chip Ganassi Racing and a former driver himself, reckons he has been utilizing electronic data acquisition systems for over 25 years now. However, the volume and sampling frequency (currently, one gigabyte per race weekend and 20 times per second, respectively) has recently been boosted by several orders of magnitude. “There’s so much now,” he points out. “What you do with the data is much more important than having it.”
Teams’ desire to constantly monitor the state of their racecar has altered the vehicle’s physical composition. With each passing season, the modern Verizon IndyCar Series car is becoming less a machine, and more like an organism that is both in symbiosis with its driver and equipped with its own nervous and circulatory systems. Its innards are packed with over a hundred sensors, a handful of onboard computers, a timing transponder and a myriad of wireless communications gear to transmit all the readings back to pit lane in real time. The stream contains an array of values that represent car velocity, engine rpm’s, steering wheel angle, gravitational forces, suspension loads, aero balance, fuel efficiency, tire pressure and temperature, just to name a few (most teams have over 1000 possible “channels” they can reference).
This data is the lifeblood of the Verizon IndyCar Series. These days, pit lane could easily be mistaken for a stock market trading floor—huddled groups of men fixated on data flowing by on a cluster of flat screens.
Just as in the markets, this wealth of data isn’t necessarily the real challenge; mustering the resources to turn it into useable information is. “The data does a really good job of showing you the cause of the problem, but it’s not really good at giving you the answer on how to solve it,” says Brian Campe, Race Engineer for the No. 2 Verizon Chevy for Penske Racing. “That’s where the driver, myself and all the engineers have to have a keen eye—a database in your head or unconscious mind to know where to look for the solution,’”
Currently, he relies on the data to give him the wide angle, or as he refers to it, “the 50,000 ft. view”, revealing all the factors at play. But it’s only valuable working in tandem with feedback from his driver, Juan Pablo Montoya. Montoya, says Campe, is the “lens that puts things into focus,” showing him where to zoom in pursuit of a solution.
Although competitors in the Verizon IndyCar Series all start with the same basic components to build their cars (a Dallara IR-12 chassis and a twin-turbocharged V6 engine), data science has become an area in which teams can openly pursue a competitive advantage. Well-heeled, multi-car teams like Penske Racing, Andretti Autosport or Chip Ganassi Racing are able pool data from all of their cars (four, sometimes five, each), giving them a much larger archive to work from than that of their competitors.
The next technology wave in the Verizon IndyCar Series looks to be a marriage of all this data with artificial intelligence, a virtual version of Campe’s “keen eye”. “What I see in the future is ability to ask the data a question like: ‘Long Beach. Turn 1. Understeer. Show me the instances when I have understeer, or the driver has said ‘understeer’, and show me what I’ve done to positively fix it and what I’ve done to negatively fix it,” Campe speculates.
An escalating technology arms race stands to widen the gap between teams who have the resources to enlist droves of human brainpower and computing brawn and those teams that don’t. However, Oliver Boisson, engineer for the one-car team KVSH Racing, doesn’t seem all that concerned. “The cars all have different aerodynamic and mechanical configurations because even though we all try to go fast in the same place, each driver is going to have a different feel…it’s a very complex system. If it were purely mechanical, a robot would be driving the car and we’d all end up with the same setup. Besides the technology, there is a massive human element to the story. It’s not necessarily putting the smartest people in the room and you will win everything, it’s finding people who can work together and understand each other. It’s an aspect of racing that is often missed, ” he says.
Boisson’s point emphasizes why the Verizon IndyCar Series is an ideal test bed for artificial intelligence. In fact, the man charged with making sure the data keeps flowing at each track, Jon Koskey, Director of Timing and Innovation for the series, admits several heavy-hitters from the tech world have approached him in recent months to express their interest in feeding the series’ race data into their AI engines.
Some are even going as far to posit that their algorithms can wrench the data towards accurately predicting who will each race.
As enticing a prospect, especially to bookmakers, as that may sound, it’s missing the point of racing—a world in which Campe reminds us “you’ll never answer a question 100% complete, there’s always a compromise.”
It’s a misconception that artificial intelligence will be fully calling the shots just a few years down the road; in fact, it’s a worst case scenario. Ideally, AI will be our attendant partner, empowering us to make better decisions. Like the relationship between driver and engineer, it first involves building trust. For example, every engineer interviewed concurs that they will happily make a decision counter to what the data says if it makes their driver more comfortable, and in turn more confident, in the car. Beyond expensive technology, it’s this nuanced understanding of context, as well as the ability to improvise and adapt, that wins races. It’s impossible to uncouple all this data from human intuition and fluidity, and that’s where there is an opportunity for us to learn how to wield the power of big data and artificial intelligence rather than be beholden to it.
Paul Biedrzycki is a New York-based writer/producer. He has been covering the automotive industry and motorsports for nearly a decade for such publications as DuJour Magazine, Esquire and Bentley Magazine. He studied English and Fine Arts at the University of Connecticut and Nova Scotia School of Art and Design.
