For Figure Skaters, AI Is Now Watching

American figure skater Andrew Torgashev recently took part in an invitation-only training camp run by U.S. Figure Skating as he prepared for the Grand Prix season. During one of his practice sessions, he landed a difficult jump that looked perfect to anyone watching in person. But the camera analyzing his movement showed he was a quarter-revolution short — and the app evaluating the jump alerted him instantly.

The tool behind that analysis is OOFSkate, an AI-powered system that evaluates a skater’s jump height, rotation speed, airtime, and landing quality without the need for sensors or wearable tech. Using a phone or tablet camera, it captures the skater’s motion, overlays ideal technique, and measures the same metrics used by technical judging panels. Coaches, judges, and skaters can immediately see whether a jump or spin was completed correctly and how the form compares to past attempts.

OOFSkate was created by Jerry Lu and Jacob Blindenbach, former college swimmers who met at the University of Virginia. Although they didn’t come from the figure-skating world, they were interested in how emerging technology could improve athletic performance and saw an opportunity to bring precise measurement tools to the ice.

The name began as a joke, inspired by the common reaction skaters have when reviewing tough feedback — “Oof!” — and later took on a second meaning: “Obsessed Over Form.”

For now, Lu and Blindenbach are moving carefully, knowing that elite sports have historically adopted new technology slowly. Their goal is to support coaches and athletes, not replace human judgment. As they put it, fully automating everything can lead to frustration and worse results; their aim is to assist, not take over.

Original Source