How AI Is Revolutionizing Olympic Figure Skating

The whirl of the Olympics often brings to mind images of athletes showcasing their prowess in their respective fields. Figure skating particularly stands out, as it merges art with athletics. With each smooth glide and gravity-bending jump, audiences can’t help but be captivated by the skater’s remarkable precision. But what’s hidden in this glittering spectacle? Well, a lot of intense physics and data-driven training efforts.

Enter, Jerry Lu, a proud MIT grad who seeks to push the boundaries of figure skating’s potential. Known for his innovative thinking, Lu has developed an advanced AI tool dubbed OOFSkate to help figure skaters land even the most difficult jumps, like a quadruple axel or the elusive quintuple leap.

The brilliance of OOFSkate and The Emergence of AI in Athletics

Thanks to his time as a researcher at the MIT Sports Lab, Lu has leveraged optical tracking and artificial intelligence to create OOFSkate. This unique system dissects video footage of a skaters’ jump and generates detailed metrics, including rotation speed and jump height. It also offers personalized suggestions for improvement. How? By translating the performance data of an athlete into actionable insights from elite skaters, a far cry from human eye analysis. “The goal is to help skaters identify areas of opportunity that might not be obvious, even to trained coaches,” declares Lu.

The artistry that we see on the ice often overshadows the complexity of the sport. Fortunately, Lu has taken up the mantle to debut this hidden intricacy. Collaborating with NBC Sports for the 2026 Winter Olympics, he plans to use AI to dissect the technical elements of figure skating. His aspirations go beyond revealing how challenging the sport is. He also aims to make the scoring system more understandable and the athletic accomplishments more relatable to viewers. “The data we collect helps reveal just how challenging it really is,” he explains.

AI’s Role in Judging and Beyond

While numbers and stats can easily quantify the technical side of figure skating, the artistic aspect provides an intriguing challenge. Professor Anette “Peko” Hosoi, co-founder and faculty director of the MIT Sports Lab, sees potential in AI’s role in offering a fair judgment on the aesthetic elements of a performance.

Backed by a grant from the MIT Human Insight Collaborative (MITHIC), Hosoi, along with Professor Arthur Bahr and grad student Eric Liu, is exploring if AI can truly “understand” beauty. Figure skating presents an excellent opportunity for such a study since it incorporates subjective criticism with concrete numerical evaluations.

One noteworthy feature of OOFSkate is its incredible use of pose estimation technology. It thrives in sports like figure skating, where key metrics — jump height, rotation, and landing precision — aren’t heavily dependent on depth. “Jerry found one of the rare sports where pose estimators shine,” Hosoi comments.

More than just enhancing athletic performance, Lu sees artificial intelligence in figure skating as a way to gain insights into human cognition and judgment. He notes that such work could have implications beyond just sports, branching out to industries where subjective evaluation plays a role such as art critique and recruitment processes.

The Road to Milan-Cortina 2026

The anticipation for Milan-Cortina 2026, the forthcoming Winter Olympics, is mounting. Lu and Hosoi are looking forward to seeing how their efforts will influence athletes and audiences alike. Being a crucial part of NBC’s data-driven storytelling for the Olympic coverage, Lu is set to help viewers appreciate each performance’s nuances and challenges.

And the ultimate challenge in figure skating, the quintuple jump? Hosoi is optimistic. “A quint is definitely possible. Not in this Olympics, maybe, but soon. A six-rotation jump? That’s probably beyond human limits. But five? That’s within reach.”

This cutting-edge blend of AI tools like OOFSkate and ongoing research into aesthetic judgment hints at a groundbreaking future for figure skating, and sports analytics in general.

Lesen Sie den Originalartikel unter MIT-Nachrichten.