Russian scientists have developed an artificial intelligence system that can detect counterfeit alcohol by its chemical "fingerprint." The technology was created by specialists from the Federal Research Center for Nutrition and Biotechnology (FSBSI).
The algorithm analyzes the composition of alcohol after laboratory testing and compares it with a database of original samples. If the parameters do not match, the system instantly signals a possible fake.
According to researchers, the neural network can detect several common falsification schemes at once. For example, when an expensive grain distillate is replaced with ordinary ethyl alcohol or foreign chemical components are added.
Modern methods of laboratory diagnostics are used for analysis - gas chromatography, capillary electrophoresis and mass spectrometry. Based on the data obtained, artificial intelligence identifies dozens of chemical markers that cannot be detected during a normal check.
Some tests take only a few minutes. This makes the technology potentially useful for laboratories and quality control systems in the alcohol market.
Such systems can become part of the digital market control and help to identify counterfeit products more quickly.