Neural Networks Monitor Food Quality in Russian Schools
In Moscow region, artificial intelligence was deployed to independently assess school meals, analyzing over 13,000 photos in the first week of the academic year to ensure menus matched official standards.
According to the Ministry of Public Administration, IT, and Communications of the Moscow region, the AI assistant was trained to detect discrepancies with the approved menu. In the first days of the school year, it processed more than 6,500 breakfast photos, 5,900 lunch images, and 500 pictures of afternoon snacks.
The method proved effective immediately. The AI controller can recognize 113 different dishes with 95 percent accuracy. If it detects an error, it instantly alerts the responsible staff member.
As a result, 619 violations were confirmed and quickly corrected. With such a strict and independent digital watchdog now monitoring children’s nutrition, officials believe that violations may soon disappear altogether.
