F1’s financier, Liberty Media, may have pushed the sport deep into the American heartland. three In the United States, Grands Prix dominate, but it’s Nascar that continues to grow its television audience while F1 and the “open wheel” IndyCar championships have declined slightly.
European race fans are notoriously apathetic towards stock car racing, but when the ostensibly low-tech, naturally aspirated, pushrod V8 engine (358 cubic inches (5.8 liters) and 670 horsepower) hurtles around an oval track, it reaches places that other racing series can’t or have only given up on.
That’s not to say the Nascar grid isn’t trying to gain technological advantages in any way possible. Lenovo is working with the series’ biggest team, Richard Childress Racing, to improve the accuracy of pit stops during races. The Nascar Cup Series has a high number of pit stops, anywhere from five to 12, depending on the circuit and on-track conditions. Notably, the company is using AI to gain real-time insights into refueling.
Fuel economy is, of course, a vital part of Nascar racing – almost an art in itself, but also a source of drama and danger. (Note: refueling has been banned in F1 since 2010 for cost and safety reasons.) Since the cars themselves are not equipped with fuel gauges in the cockpits, it is the responsibility of the team strategists to constantly monitor how much fuel is being added during pit stops, and how quickly it is being consumed.
As with any use case, fuel consumption depends on many variables, including the length and configuration of the track, the speed the cars are traveling, etc. There are several “cautions” during the race where the cars will typically use half the fuel.
In NASCAR, drivers also use a technique called “drafting,” which allows them to maintain speed in the pack without using full throttle. Less fuel consumption means fewer pit stops, and when they do, they use less fuel. On average, NASCAR Cup Series cars burn about 100 gallons (380 liters) of fuel per race, even though they’re not the most energy-efficient.
Lighter is always faster
It’s not an exact science, but the goal of Lenovo’s AI team is to make it as accurate as possible: If the RCR could measure how long the fuel can was plugged into the car, the team figured it could more accurately calculate how much fuel was being delivered.
That was the brief. Lenovo’s answer was to devise a system that would use on-board transponders and a camera mounted above RCR’s pit box to identify when a car entered the box and initiate a real-time video feed.
“An AI engine looks at each frame and classifies whether a fuel can is plugged in or not,” explains Lenovo AI data scientist Sachin Wani. “We’re working at 30 frames per second, so the information is accurate to within about 0.03 seconds. Previously, fuelers knew they would have to pump about seven seconds’ worth of fuel without an auxiliary device due to safety concerns.”
“So you’re basically just doing the maths in your head and seven seconds could have become eight or nine or, even worse, five or six. That obviously throws off your strategy and creates a situation where you run out of fuel and have to make another pit stop,” Wani said.
