The functioning of the semi-automatic offside is about to change radically in view of the next 2026 World Cup, thanks to an unprecedented technological evolution that will retire the old and much discussed two-dimensional reconstructions. FIFA will introduce for the first time a 3D VAR system which, thanks to realistic three-dimensional avatars of the 1248 players of the 48 national teams, also generated with the help of artificial intelligence, will faithfully reproduce every player on the pitch thanks to scans taken in special environments before the start of the championship.
How the 3D VAR of the 2026 World Cup works
Developed together with technological partner Lenovo, the new technology adopted by FIFA abandons standardized silhouettes and flat lines to rely on personalized digital models, built from real body scans of each of the over 1,248 athletes participating in the 2026 FIFA World Cup, the first edition to be held in three different nations: Canada, USA and Mexico.
Thanks to generative artificial intelligence and a sophisticated hardware infrastructure, the system will track every part of the body with millimeter precision, producing fluid and visually accurate 3D animations, broadcast both in VAR rooms and on televisions around the world. The objective is twofold: on the one hand to support referees in making decisions very quickly; on the other hand, eliminate the controversies regarding the perspective and transparency of replays.
To understand the scope of the change, it is worth remembering where we start from. The SAOT system (Semi-Automated Offside Technology) is now based on a network of 12 special cameras installed under the roof of the stadiums, capable of tracking 29 nodal points of each player’s body: knees, shoulders, ankles, etc. The exact moment of the pass is isolated thanks to a sensor inside the ball, which records stress data 500 times per second. The Achilles’ heel of the system, however, is the visualization of the actions: the geometric lines superimposed on a flat broadcast image can generate inevitable perspective distortions and, precisely these, fuel the skepticism of spectators and fans.
With 3D VAR this should no longer happen. Before the start of the tournament, during the official Media Days, each athlete will enter special scanning stations (like the ones you see in the following image).
A group of 36 connected cameras will capture geometric proportions and textures in less than a second. All in 30 seconds. The volumetric segmentation algorithms will then isolate the body mass in space, returning a raw mesh avatar: a network of three-dimensional polygons that faithfully replicates height, shoulder width and limb length, with a tolerance of less than half a centimeter.
In case of potential offside, the data processed by Hawk-Eye Innovations will no longer animate generic mannequins, but precise digital clones, immersed in a virtual space freely explorable by the camera. If the offside exceeds 10 centimetres, the line assistant will immediately receive a signal on their device and will be able to raise the flag without having to wait for the end of the action.
The technology that calculates offside is made up of over 17,000 devices
The machine that supports all this is impressive: over 17,000 Lenovo and Motorola devices involved across servers, data centers and local units distributed in the sixteen stadiums hosting the competition. But the amount of data collected will not only be used for VAR. FIFA will also launch Football AI Pro, a virtual assistant based on a football language model that will allow all 48 teams to analyze over 2,000 athletic and tactical metrics before and after each match, making advanced Match Analysis accessible to anyone. And even the Referee Camera, the camera worn by the referees, will benefit from real-time stabilization algorithms, eliminating the blur of the race and offering spectators a clear and immersive first-person view.
