Listen to the article
In a breakthrough for figure skating technology, a new AI-powered application called OOFSkate is revolutionizing how skaters train and could eventually transform how competitions are judged. The system, which requires nothing more than a tablet or mobile phone, provides instant, precise feedback on jumps, spins, and other technical elements that traditionally rely on subjective human evaluation.
American figure skater Andrew Torgashev recently experienced the system’s precision firsthand at an invitation-only U.S. Figure Skating camp. While attempting a quad toe loop—a four-revolution jump launched from the back outside edge of the blade—Torgashev appeared to nail the landing perfectly to observers. However, the OOFSkate app immediately detected that his jump was a quarter-revolution short, a minute error that could prove costly in competition where fractions of points determine medal standings.
“Our vision for the system is to automate the technical calling of the sport,” explained Jerry Lu, who developed the technology with his former college roommate Jacob Blindenbach. “We’re using AI-assisted computer vision combined with figure skating knowledge to remove elements that shouldn’t be judged subjectively.”
The concept represents a fundamental shift in approach: let human judges evaluate the artistic components while computers handle the technical measurements with precision and consistency.
The system operates with remarkable simplicity. Using a standard camera on a mobile device, it captures a skater’s movements in real-time, overlaying key points of jumps or spins against an idealized version of each element. It instantly records crucial metrics that technical panels use—rotation completeness, correct edge usage, jump height, and spin speed.
For coaches and skaters, this means immediate, objective feedback during training sessions. Athletes can compare their current performance against their previous attempts or even benchmark against elite skaters in a team library. “I can do a quad toe loop and compare it against myself, along with all the other athletes that have executed this,” Lu explained, using Torgashev as an example. “Am I spinning as well as Mikhail Shaidorov? Or a previous version of myself?”
Lu and Blindenbach, both former collegiate swimmers from the University of Virginia, weren’t initially focused on figure skating. After graduation, Lu pursued studies at the MIT Sports Lab while Blindenbach specialized in artificial intelligence at Columbia University. Their shared interest in using technology to enhance Olympic sports kept them connected.
It was NBC, the U.S. Olympic broadcaster, that ultimately steered them toward figure skating. The network sought technology to help commentators Tara Lipinski and Johnny Weir provide more detailed real-time analysis during broadcasts. U.S. Figure Skating quickly recognized the training potential and joined the project, with Olympic skaters Jason Brown and Alysa Liu, along with coach Massimo Scali, providing valuable feedback.
The developers regularly visit the prestigious Skating Club of Boston to test and refine their system—potentially initiating a technological revolution in a sport that has traditionally been slow to embrace change.
The name OOFSkate originated from skaters’ typical reaction to seeing their jump feedback—”Oof, that wasn’t very good!” Later, someone at U.S. Figure Skating repurposed it as “Obsessed over form,” reflecting the technology’s meticulous attention to technical details.
The system’s ability to eliminate subjectivity from technical evaluation parallels similar technologies that have transformed officiating in tennis and may soon revolutionize baseball’s strike zone. “If someone under-rotates,” Blindenbach noted, “that should always be called. There shouldn’t be a missed call or controversy. Sometimes a judge’s position makes it hard to see if a skater is on the correct edge for a lutz or flip. We hope AI can make the sport more fair.”
Despite the technology’s promise, the developers remain realistic about its implementation timeline. They point to Hawk-Eye in tennis, which took nearly two decades before fully replacing human line judges at Wimbledon. Olympic adoption faces additional hurdles, including existing partnerships with official data provider Omega.
For now, Lu and Blindenbach are concentrating on refining OOFSkate to support coaches, athletes, and commentators as they prepare for the 2026 Milan Cortina Winter Olympics.
“We don’t want to step on toes,” Blindenbach emphasized. “When you go fully AI and remove humans from the process, people generally resist. We want to assist with objective elements like jump height, rotation, or under-rotations, which AI handles well. The artistic aspects remain where human judgment excels—that’s where we see our role.”
Fact Checker
Verify the accuracy of this article using The Disinformation Commission analysis and real-time sources.
7 Comments
Fascinating to see how AI is being applied to improve figure skating training and judging. Objective, precise technical feedback from AI could really help skaters hone their skills and eliminate subjective scoring issues. I’m curious to see how this tech is adopted across the sport.
While the potential benefits of AI-powered figure skating judging are clear, I imagine there will be some resistance from traditionalists in the sport. Relying solely on computer vision to evaluate technical elements may raise concerns about losing the ‘human element’ in competitive figure skating. Striking the right balance will be key.
The development of AI-assisted figure skating judging is an intriguing step forward, but I wonder about potential downsides. Could over-reliance on technology diminish the artistry and expression that are integral to the sport? Careful implementation will be crucial to maintain the sport’s essence while improving technical evaluation.
That’s a very valid concern. The challenge will be striking the right balance between objective technical assessment and preserving the artistic, emotive qualities that make figure skating so compelling to watch. Thoughtful integration of the AI system will be key.
Removing human bias and error from technical scoring in figure skating seems like a big step forward. The OOFSkate app’s ability to detect even small deviations could make a real difference in high-stakes competitions. I wonder how skaters and coaches will incorporate this kind of AI feedback into their training.
You raise a good point. Integrating real-time AI analysis into figure skating training could allow skaters to fine-tune their technique in ways that were previously very difficult. It will be interesting to see how this impacts performance at the elite level.
As someone who enjoys watching figure skating, I’m excited about the prospect of more objective, consistent scoring through AI technology like OOFSkate. Eliminating subjective bias could really enhance the integrity and fairness of the sport. It will be fascinating to see how this evolves in the lead-up to the next Olympics.