Students Help Develop an AI System to Support Sports Referees
SberAI, part of Sberbank, is developing new technologies with students. A judging-assistance system built through this collaboration has already been tested at major national-level competitions.
Students at the School X Institute of Advanced Technologies at Don State Technical University (DSTU), working under the guidance of SberAI specialists, have developed an artificial intelligence system to support sports judging. The technology has already been used at the Magnit Russian Weightlifting Championships.
A Helping Hand for Referees
The AI judging-assistance system uses computer vision to analyze video of an athlete’s performance, tracking movement and calculating how the lift is executed. AI helps turn criteria that judges traditionally assess by eye into measurable indicators. The system evaluates limb positions, movement trajectories, the stability of the final position and specific signs of rule violations.
Naturally, until now judges have simply watched each lift and assessed the same parameters themselves, without any such system. But judging by eye can’t guarantee perfect accuracy. Judges don’t need someone to make decisions for them; they need precise information. That is what this new technology is designed to provide.
The developers emphasize that AI does not make decisions on behalf of judges. Instead, it supplies data to help them make objective assessments, particularly when a call is disputed.
Hands-On Experience Before Graduation
Fourth-year students Alena Kuryatnikova, a product manager, and Vladislav Petlyuk, a developer, worked on the project alongside the SberAI team. They had the chance to move beyond theory and gain practical experience, not just by developing conventional products, but by working with industry professionals to build an entirely new system.
“Working on this project with the SberAI team was a valuable experience for us. We handled the technical side of the solution and worked with experts to understand the rules of weightlifting and translate them into evaluation criteria for the AI system. What’s especially exciting is that our work has gone beyond development and is now being used in real competitions at the national level,” said Vladislav Petlyuk, a student at DSTU’s School X Institute of Advanced Technologies.
AI is only beginning to make its way into sports judging. But demand for these technologies is already strong enough that the Russian Weightlifting Federation tested SberAI’s system at national-level competitions. The results of those trials also convinced federation leaders that the system should continue to be used in judging.
Technology for World-Class Sports
Using AI to analyze athletic performances is part of a global trend. Russian systems are not just in development; they are already being used in practice and becoming routine tools for sports analytics. Russia is also building a dedicated SportTech market, encompassing systems that evaluate athletes’ performance, analyze movement, automate judging and provide digital support for training. Russian technologies that have demonstrated their effectiveness in real-world use could eventually be adopted in other countries that are digitizing sports but lack comparable homegrown solutions.
SberAI experts also highlighted the students’ role in the project, describing this kind of hands-on work as an essential step in professional training. Having a project like this in their portfolios could help the students land jobs and advance their careers by taking part in even larger and more complex AI development projects.
