A UTK engineering team is building a computer vision system that watches football players during practice and games, detects flaws in their movement, and flags athletes who may be at higher risk of injury — before they get hurt.

The project is led by Assistant Professor Hector Santos-Villalobos and the MARCI (Multi-Modal Analytics, Reasoning, and Computational Imaging) team in UTK's EECS department. The system uses computer vision to detect players in video footage and estimate key body positions — head, wrists, ankles, knees, and hips — to identify dangerous movement patterns. It was awarded $60,000 in seed funding through UTK's AI TechX industry consortium in September 2025. Partners include UT Athletics' Director of Applied Performance Science and the Joe Gibbs Human Performance Institute in Charlotte, NC. The project is targeted for completion by the end of summer 2026, with potential expansion to other sports.