Russian Scientists Develop AI System to Detect Dangerous Asteroids

A new Russian AI-powered system promises to identify hazardous celestial bodies and improve early detection of space threats.

Researchers from Moscow Institute of Physics and Technology (MIPT) have developed a cutting-edge system to analyze and detect hazardous comets and asteroids. The project, known as 'Strazh' (Guardian), was reported by Argumenty i Fakty citing Vitaly Shelest, the project’s scientific supervisor at the PUSK Center at MIPT.

The system runs on Russia’s own Yandex Cloud infrastructure and processes telescope data—an inherently complex task due to the enormous volume of information generated nightly. Globally, telescopes can produce up to 15 terabytes of data per night and that ammount is equivalent to roughly 30,000 hours of high-resolution video.

After initial analysis, Strazh classifies stars that exhibit changes in visible brightness. Its classification accuracy currently reaches 87 percent.

While the system's main goal is to detect potential cosmic threats earlier it is already capable of discovering new stars and supernovae.