Working with Lenovo, Richard Childress Racing (RCR) uses AI at the edge to gain real-time insights into fueling during pit stops. By understanding exactly how much fuel is loaded during each stop, the organization can time future stops strategically and maximize its chances of raceday success.
Founded in 1969, RCR is one of the largest racing organizations in NASCAR. Competing in the NASCAR Cup Series and the NASCAR Xfinity Series, RCR has built up a strong reputation for excellence, innovation, and a winning spirit.
For RCR, success on the track is a true team effort. While the skill and expertise of its drivers is crucial, control-center teams must also make effective strategic decisions, including the timing of pit stops during races.
Every second counts
In NASCAR events, fueling is an art. Each car must be fueled using a specially designed fuel can that enables pitcrews to rapidly dispense up to 11 gallons of gas at a time. RCR recognized that optimizing its approach to fueling was a big opportunity to save valuable seconds during races. The aim was to calculate the exact fuel requirements for each car and strategically time its fuel stops to maximize opportunities for success on the track.
To achieve this goal, the organization needed to know how much fuel was loaded during each fuel stop. If it could measure the amount of time that fuel cans were connected to its cars, RCR could calculate the quantity of fuel delivered—but this was no easy task in the fast-paced environment of the pitbox.
Engaging a trusted partner
Over the last three years, RCR has been partnering with the Lenovo OEM Solutions team to implement the latest technology and supercharge its operations with data-driven insights. Working closely with the OEM and engineering teams to identify ways to achieve better outcomes on the track, RCR proposed developing a solution to optimize pitstops by capturing fuel data using the power of AI.
Lenovo developed an innovative solution, exclusively with RCR, to capture fueling data. Using in-car transponders and a camera installed overhead the pitbox, the solution identifies when an RCR car has entered the box and starts a real-time video feed. An advanced AI engine then processes 30 frames of video per second to detect the precise times that a fuel can is connected and disconnected from each car.
The solution is hosted on a Lenovo ThinkEdge SE360 V2 server to which multiple requests can be made by anyone authorized on the network at the track. It is hosted as an API (application program interface), which makes it very easy to use.
Sachin Gopal Wani, AI Data Scientist at Lenovo’s AI Discover Center of Excellence, comments: “Delivering a solution that is easily interpretable by the user is a crucial part of our AI leadership. The API provides a browser viewpoint for anyone to check the status of pitstops and the duration, along with the log. The real-time results can be auto-populated in Excel for race engineers, supporting strategic decision-making.”
High performance at the edge
With the Lenovo ThinkEdge server equipped with an NVIDIA® L4 Tensor Core GPU, RCR can drive high-performance for AI workloads trackside. The inferencing automatically starts when the car enters the pitlane and stops when it exits the pitlane. The solution can handle multiple cars coming in at once. In fact, it’s so efficient that, mathematically, it can handle 30 cars at once in real time. The solution measures fueling with 99.4% accuracy in real time, even at night or in poor weather—helping RCR gather deeper insights.
RCR is already putting this new source of analytics insights to valuable use. The organization uses the AI solution to power a digital countdown display for its drivers and crew—letting them see the progress of fueling in real time.
Most importantly, AI at the edge is helping RCR to make better-informed strategic decisions about when to load fuel during races. Thanks to data insights powered by Lenovo, the organization can gain up to four positions per pit stop and sharpen its competitive edge. Since knowing the fuel level is crucial, and pitstops can make or break races, the solution will equip RCR with the knowledge to build strategies and potentially win future races.
Racing ahead with Lenovo
RCR’s work with AI at the edge is just getting started. To further optimize refueling, the company is working with Lenovo to calculate the angle of the fuel plug to gain an even clearer picture of the amount of fuel in each car. RCR and Lenovo are also exploring an innovative augmented reality (AR) solution that projects a display timer in front of the fuelman that displays the amount of time that the fuel can has been plugged in.
Next, RCR plans to work with Lenovo to develop an AI solution to help streamline tire-changes—another crucial pit-stop activity that plays a big role in race-day success. This includes using audio and visual feed to determine whether all the lug nuts on a car’s tires are adequately tightened during a pitstop.
To learn more about Lenovo’s work with RCR, click here. To build an exciting custom solution for your use case, contact us.
