Formula One sponsor Liberty Media may have taken the sport deep into its American heartland: there are now three Grand Prix in the USA, but it is Nascar which continues to increase its television audience there in contrast to a slight decline in the “open wheel” championships of F1 and IndyCar.
European racing fans are famous for their disdain for stock car racing, but there’s something about a seemingly low-tech, pushrod-powered, naturally aspirated V8 (with a capacity of 358 cubic inches (5.8 liters) and output of 670 horsepower) racing around an oval and going places other racing series can’t (or perhaps have stopped) reaching.
It’s not that the Nascar grid isn’t trying to gain a technological advantage by any means possible. Lenovo is working with one of the biggest names in the series, Richard Childress Careersto help fine-tune its pit stops during a race, and there are plenty of them in the Nascar Cup Series – anywhere from five to 12, depending on the circuit and what’s happening on the track. In particular, the company is using AI to gain real-time insights into refueling.
Fuel consumption is obviously a fundamental part of any Nascar race, almost an art in itself, as well as being a source of drama and danger. (Note: refuelling has been banned in F1 since 2010 for reasons of cost and safety.) The cars themselves are not equipped with fuel gauges in the cockpit, so team strategists must constantly monitor how much is coming in during a pit stop and the rate at which it is being consumed.
As with any other use case, fuel consumption depends on a number of variables, including track length and configuration and the speeds at which the cars are travelling. There are a number of “cautions” during a race, when cars typically burn half their fuel.
In Nascar, drivers also “draft,” a technique that allows them to maintain speed in the pack without using full throttle. Lower fuel consumption means fewer pit stops, and when they do, they consume less fuel. On average, a Nascar Cup Series car (not the most energy-efficient device) will use 100 gallons (380 liters) of fuel in a race.
Lighter is always faster
It’s not an exact science, but the goal of Lenovo’s AI team is to get as close to it as possible. If RCR could measure the amount of time the fuel canisters were attached to their cars, they reasoned, then the team could more accurately calculate the amount of fuel delivered.
That was the idea. Lenovo’s response was to design a system that used transponders on the car and a camera mounted above the RCR pit to identify when a car had entered the pit and begin streaming real-time video.
“An AI engine analyses each frame and classifies whether the fuel tank is capped or uncapped,” explains Sachin Wani, AI data scientist at Lenovo. “We work at 30 frames per second, so the information is accurate to within about 0.03 seconds. Before this, the fuel attendant knew he had to fill up with fuel for about seven seconds, without any safety aids.”
“Basically it came down to mental calculations, which meant that seven seconds could become eight or nine. Or worse, five or six. That obviously messes up the strategy and creates a situation where you run out of fuel and need to make another pit stop,” says Wani.