The Premier League kicks off this week and a new AI refereeing tool is coming to help spot when players are offside before the competition ends. reported by Wired. The league has turned to sports tech company Genius Sports and its Semi-Assisted Offside Technology (SAOT).
Unlike typical video assistant referee (VAR) systems, SAOT doesn’t require expensive 4K cameras, just a network of 24 to 28 iPhones positioned around the pitch. The on-board GeniusIQ AI then analyses the footage gathered by the iPhones to work out exactly where every piece of a player was at any given moment, and whether they were in an offside position.
The rules of sport can be complicated, and few are as notorious as offside fouls for being so contentious and difficult to call. VAR has helped, but is often criticized for slowing down the game or leading to wrong decisions. SAOT can look at the pitch and collect between 7,000 and 10,000 data points per game, creating detailed 3D virtual meshes of each player. By comparison, current systems often rely on 30 to 40 tracking points to create rudimentary stick figures for players,
The GeniusIQ system’s 3D meshes provide a much more detailed and accurate model. That data then creates a picture of where each player is in relation to the ball and the offside line. iPhones, with their 100fps capture rate, are more than fast enough to run the GeniusIQ system. The AI, trained on multiple seasons of games, can identify not only player positions but individual body parts as small as their fingers, and extrapolate even if they’re partially obscured.
AI refs
The Premier League’s adoption of this technology could streamline the decision-making process, reducing the delays that have plagued VAR use in the past. While an exact rollout date has not yet been announced, it is expected that the technology will be fully in use at all Premier League matches soon.
However, while the SAOT system has shown promise in testing, its effectiveness in live matches remains to be seen. It will need to demonstrate that it can consistently outperform existing VAR systems in terms of accuracy and efficiency. Furthermore, it will need to overcome a reliance on extensive data processing, meaning that technical issues with the servers or software could cause the system to fail.
As the SAOT system is implemented in live matches, it will be watched closely by the global football community. If successful, it could set a new standard for the use of technology in sport, not just football but a range of athletic sports where the precise location of the match can determine the fate of a match. Ultimately, the human referees make the final decision and are either celebrated or cursed, regardless of what the AI suggests about who is offside.