Russia Boosts Computer Vision Algorithms for Mining Industry
A new AI system at Severstal’s Yakovlevsky mine is cutting downtime by spotting clogs and overloads in real time, saving millions of rubles.
The Yakovlevsky Mining and Processing Plant, part of Russian steel giant Severstal, has deployed a computer vision model that eliminates delays in monitoring ore transport underground. Previously, dispatchers had to manually track ore movement, often identifying problems too late.
The new algorithm connects directly to video streams, automatically detecting clogs or overflows and sending operators a photo report and alert through the “Steel Vision” monitoring system.
Developed by Severstal Digital in collaboration with the plant’s staff and the Severstal-Infocom Center, the solution has already prevented 19.3 hours of downtime in just six months. Experts estimate the economic benefit at 2.9 million rubles (about $32,000).
This is not the first AI application in Russia’s mining sector. Earlier projects included a virtual assistant designed to process field data.
