Technical Partner — ML, XAI & User Interface
We built the machine learning and explainable AI models that estimate athletes' injury risk, and the user interface through which staff explore the data and predictions.
Musculoskeletal injuries in elite team sports are rising as the number and intensity of matches and training sessions grow. They strain health systems, shorten athletes' careers and cost clubs heavily. Even with regular health checks, medical staff can't reliably tell which athletes are most at risk.
IRONMAN monitored four football and four basketball teams, men's and women's, through a full competitive season, using GPS, accelerometry and inertial sensors. It assessed each athlete's physical and functional condition at three points across the season and recorded every injury. We turned the data into injury-risk predictions using machine learning, data mining and explainable AI, presented through a decision-support interface.
