In 2024, the NFL had its lowest concussion rate on record, decreasing 17% from the previous year, thanks to the NFL’s AI-powered athlete performance and injury prevention technology, the Digital Athlete. After years of concussion research and partnerships, Jennifer Langton, former senior vice president of player health and innovation for the league, spearheaded the AI-led approach to injury prevention.
Langton was a college athlete whose career ended abruptly with a knee injury. “One of my greatest challenges in my life became inspiration for transformation to help to mitigate [injuries] from happening [to] NFL players,” she says.
Langton joined the NFL after a stint as the North American CFO of Atari. After graduating with her MBA, she carved a new position in the NFL’s health and safety division, tackling the concussion problem.
A technological breakthrough
Work on the Digital Athlete began years ago, with a series of strategic partnerships. “We spearheaded with GE [General Electric]. What we called it was a $60 million head health initiative… to better understand diagnostics for brain injury,” Langton explains. This early collaboration laid crucial groundwork by bringing together diverse expertise. “The foundation of that work for me was the breakthroughs that we had when you collaborate with others rather than [doing] it on your own.”
This foundational research on head health evolved into a systematic analysis of how injuries happen on the football field with Jeff Crandall, Ph.D., an expert in auto safety. “His insights allowed us to have a new perspective on how he would take a car crash [and] simulate it in the lab… to mitigate the forces and impacts that lead to injury,” Langton explains. The team adapted Crandall’s crash-test methodologies specifically for athletes, applying automotive safety engineering principles to the unique dynamics of football collisions.
Initially, the NFL’s approach to understanding injuries involved significant manual effort. “Dr. Crandall and his team would review every single concussion by human eye, labeling 150 variables, defining what happened in that concussion,” Langton says. “To count the number of head impacts per game took four days.” This tedious process limited how quickly teams could apply insights.
The turning point and idea to use AI to process the data faster came during a visit to GE’s innovation lab, where Langton saw a “digital twin” used to predict failures in turbine engines. A digital twin is a virtual recreation of a real person or object used to simulate how that person or object would react under physically stressful conditions. “When we left, we said, ‘Well, what if we create the digital athlete?’” she recalls.
To bring this vision to life, the NFL did an extensive road show tour to find the right technology partners. “We visited everybody from Microsoft, Google, Amazon Web Services,” she says. They chose AWS because “they were strategically aligned to [the NFL’s] vision. If we wanted to predict injuries, we knew that there were technologies that didn’t exist that we’d have to build. And they saw our vision and agreed to it. Our end goal was to reduce injuries and optimize performance.”
Langton’s team also visited academic institutions like “MIT and Columbia, because they knew where the R & D was for new technologies like pose estimation—how do you quantify the body movement and computer vision? Because it was just at its infancy then,” she says. With the help of technology and AI-powered scientific research, Langton’s team created the Digital Athlete.
How the Digital Athlete works
The Digital Athlete platform creates a virtual representation of every NFL player by integrating multiple data sources, like game and practice video footage, sensors in shoulder pads that track player movements, performance metrics from team training, and biomechanical analysis of impacts.
With computer vision technology, the system analyzes what once took days in seconds. The system identifies specific movement patterns that lead to injuries for individual players or positions. For example, “if a quarterback is rushed, he braces the ball and not himself. And when that happens, he is more likely to get injured in the back of the head… because he’s holding onto the ball,” Langton explains. This insight allowed teams to modify training techniques and equipment design to address specific injury risks.
The data gathered from the Digital Athlete platform transformed helmet design. Crandall’s theory was, “Why isn’t a quarterback’s helmet designed with more energy-absorbent material in the back of the head to prevent those forces from impacting that position?” Langton says. “A lineman takes different frontal lobe hits. Why isn’t a helmet designed to prevent that impact from happening to those players?”
By sharing their findings with manufacturers, the NFL enabled the creation of position-specific helmets designed to address the unique impacts experienced by different players. The results were dramatic: “We had nine times the rate of innovation for helmet performance. It was unheard of,” Langton says.
Equipment managers for each team helped sell the idea of these new helmets to players. “We started that education one-on-one with every team. One year, we [had] 40% of players in top-performing helmets that tested well in the lab. The next year, we went to 99%,” Langton says.
People at the center of technology
Humans were at the center of every step of the Digital Athlete’s development process. “The 32 NFL teams have the most exceptional athlete management teams…. We [had] to put them at the center. We [had] to say, ‘What is it that you’re doing, and what data do you need to do your job better?’” Langton says. “Why? Because they know the athlete. They see the athlete every day.” The NFL focused on making the technology serve the people using it rather than forcing people to adapt to the technology.
The NFL can now integrate and aggregate data across all 32 teams. “If you think about one running back, one team only manages a 53-man roster. But if you can compare that running back across all running backs across the league, that’s the power of what we were bringing to them,” Langton says. The Digital Athlete can spot injury risks for specific players and work across teams to find patterns between particular positions.
Langton sees AI not as a replacement for human creativity and productivity but as an enhancement of it. “AI is more of an empowerment tool, not a replacement, to help [people] do their jobs more effectively and efficiently,” she says.
The NFL’s success shows that when organizations partner and keep the humans who will use the technology at the center of the process, they can transform how people work, whether on the football field or in the office.
Photo by Danny E Hooks/Shutterstock
