Russian AI Boosts Mining Efficiency
A mill is a massive rotating drum filled with chunks of ore and steel balls. As the drum turns, the balls grind the ore into sand. The material is then washed and processed to separate valuable minerals from waste rock. But ore is far from uniform. The amount of valuable minerals recovered depends on the hardness of the feed, the drum’s rotational speed and the volume of water used.
Operators must keep track of dozens of parameters, a demanding task that makes it difficult to respond quickly to changing conditions. As a result, mills often operate outside their optimal range: they may be overloaded, underloaded or consuming excessive amounts of electricity. ASTECH Industries has announced plans to deploy an intelligent grinding controller at a Russian mining and processing plant.
A Mill on Autopilot
The company has developed an “autopilot” for grinding mills designed to keep them operating at optimal settings. Sensors continuously measure what is happening inside the equipment: how much ore it contains, how the load is distributed, how hard the steel balls strike the drum walls, how much water is flowing in, the level of vibration and the motor’s power consumption.
The data feed into software that compares actual operating conditions with an ideal target. If the ore becomes harder, the system determines that the feed rate should be reduced slightly and more water added. If the material is softer, it can increase the feed rate and speed up the rotation.
he controller sends commands directly to the equipment and immediately checks the results. If performance improves, it retains the adjustment; if not, it makes further changes. Unlike conventional automation, the intelligent controller does not rely on a preset response for every situation. It analyzes the data and determines the best course of action.
A dedicated system monitors the mill’s condition. Vibration sensors detect how the steel balls strike the mill lining, the protective layer inside the drum. The pattern of these impacts can indicate how full the mill is, whether the balls are hitting one area too hard and whether the equipment may be at risk of damage.
Smart Technology Gains Ground in Mining
Similar technologies for intelligent grinding control are already in use across Russia’s mining sector, including systems developed by ASTECH Industries.
For example, Norilsk Nickel began using ore-crushing optimization algorithms at its Bystrinsky Mining and Processing Plant five years ago. The system tracks hundreds of operating indicators from measurement equipment in real time, cutting the time needed to detect abnormal readings by 30%. It was later upgraded to do more than recommend adjustments: it could control the process itself. Mill throughput increased by 2.64%.
Ore can vary significantly across different parts of a deposit, so efficient processing requires tailored approaches. EuroChem therefore introduced an application called Mineralogy, which uses computer vision and deep learning to analyze ore quality directly at the mine site, without requiring specialists to perform manual checks. The system allows operators to adjust processing plant settings in time to improve mineral recovery. The industrial solution, reportedly the first of its kind in the world, has already improved ore quality and reduced waste-rock content. It has also cut the share of geologists’ time spent on routine tasks from 30% to 6%.
Using intelligent algorithms, the Kachkanar Vanadium Mining and Processing Plant, part of the EVRAZ Group, identifies ore quality in each excavator bucket and directs the material to designated stockpile areas. This makes it easier to select the optimal operating settings for the processing plant. The digital approach saves the company 139 million rubles a year (approximately $1.7 million).
Running at Peak Efficiency
Keeping a mill at its optimal operating point is critical to its efficiency. More ore can be processed in the same amount of time, without building a new plant. Grinding mills are among the largest energy consumers at a mining and processing operation, so even a reduction of a few percentage points in electricity use can save millions of rubles a year. Better control can also mean fewer equipment failures and less downtime. The monitoring system can flag potential problems before equipment breaks down.
According to the developers, combining the controller with the monitoring system has already delivered measurable results. At some operations, the technology increased feed stability by 22% to 31%, raised the feed rate of raw material by 1.3% to 4.2% and reduced specific energy consumption by 2% to 8.2%.
Further pilot deployments are planned at Russian mining and processing plants, followed by a transition to routine operation. Over time, the technology could be extended to flotation and other processing stages, creating integrated production control systems. Export opportunities are also emerging. This year, ASTECH Industries reported deploying grinding and flotation control systems at Iran’s Sungun copper mine. The systems reduced energy consumption by 11% and increased productivity by 6.6%. Russian digital technologies for mining have therefore demonstrated their effectiveness both at home and abroad.
