Russian Mining Plant Deploys AI Vision System to Optimize Ore Processing

A Murmansk-based mining facility is using neural networks to scan raw ore in real time, reducing downtime and preventing damage to expensive equipment.

The Kovdor Mining and Processing Plant in Russia’s Murmansk region has implemented the ML Sense machine vision system to monitor the quality of ore flow and mitigate operational risks. This AI-powered assistant continuously inspects the conveyor belt, detecting non-crushable inclusions in real time.

The neural network is capable of spotting foreign objects as small as 10 mm, including bits of metal, wood, rubber, plastic, and other materials that could damage industrial machinery. When a hazardous object is detected, the system immediately halts the conveyor. Its detection accuracy is rated at 90 percent, with a response time of less than one second.

By preventing unscheduled downtime caused by debris, ML Sense improves the recovery rate of valuable materials and eliminates the risk of equipment failure. The project is part of a broader national strategy to digitize Russia’s mining sector.