Neural Networks Move Into Metallurgy
At the pelletizing and metallization plant of Oskol Electrometallurgical Combine (OEMK), engineers have completed the first phase of deploying industrial artificial intelligence to optimize operations of a metallization unit – a core system that produces metallized pellets used alongside hot briquetted iron in steelmaking. The platform processes large volumes of operational data and suggests operating settings to plant engineers.
Pilot results show the system can keep iron and carbon levels within target ranges in pellets. Meanwhile, it continues to train on new data, steadily improving the precision of its recommendations.
OEMK, part of Metalloinvest Group, ranks among the most technologically advanced steelmakers. It is the only full-cycle metallurgical facility in Russia that combines direct reduced iron production with electric arc furnace steelmaking. If the solution shows a measurable economic impact, the company plans to roll it out across other production units and potentially expand it across Metalloinvest’s broader asset base.
The Era of Smart Manufacturing
The OEMK case signals a broader shift toward AI-driven manufacturing. In practice, algorithms integrated with industrial IoT and cloud platforms are changing how production runs and optimizing plant performance.
These solutions could find export demand, even if indirectly. Countries expanding DRI/EAF steelmaking routes – where operators need to modernize existing facilities quickly – are increasingly looking for operating control models, situational center architectures, data preparation algorithms, and integration with SCADA and MES systems. In this context, Metalloinvest, as a leading global supplier of hot briquetted iron and iron ore products, could help bring these solutions to international markets.
More broadly, industrial AI is becoming a major factor in improving efficiency across Russian manufacturing. Today, 46% of Russian plants use AI to control quality and predict equipment failures. The impact is measurable: between 2023 and 2024, the technology added 500 billion rubles (about $5.5 billion) in profit to the industry, while reducing accident rates and operating costs by 15%. Analysts expect the Russian AI market to exceed $40 billion by 2030, with annual growth of 26.5%.
Industrial Giants Bet on Algorithms
In 2025, RUSAL, one of the world’s largest aluminum producers, began using AI to run microstructure analysis of aluminum ingots. That shift improved product quality and extended the lifespan of pressing equipment, reducing analysis time per sample from several hours to just 15 minutes. In April 2026, the company announced a neural-network-based technology to manage aluminum hydroxide particle size distribution. Now in industrial use, it has reduced fine fractions in finished products by 4.4%, which improves alumina quality, enhances electrolysis performance, and lowers energy consumption per ton of aluminum.
Nornickel ranks among the global top three in AI adoption in mining and metallurgy, alongside Tata Steel and POSCO. The company deploys intelligent systems to manage critical production processes. In ore processing, neural networks now control 70% of equipment, while real-time algorithms optimize grinding mills, flotation systems, and Vanyukov furnaces. Since 2021, AI now supports the full production cycle, from ore crushing to smelting. At the Talnakh concentrator, three AI agents coordinate maintenance operations. The company is also developing platinum and palladium alloys designed to replace gold in electronics. The company estimates that AI adds tens of billions of rubles in value. By 2030, Nornickel expects AI-driven gains to reach about 50 billion rubles (around $550 million).
Industrial AI Becomes a Strategic Asset
In 2025, Magnitogorsk Iron and Steel Works (MMK) set up an AI center of expertise to develop and deploy advanced technologies to improve operational efficiency. AI is already widely used at MMK to process large datasets, optimize production, reduce costs, control product quality, and improve workplace safety. Over the past five years, more than 50% of digitalization projects at the company have incorporated AI.
The new center will bring together expertise in machine learning, computer vision, and large language models, creating solutions that can scale across the broader industrial sector.
Together, these developments show that industrial AI in Russia has moved beyond isolated pilot projects. Leading companies are making it part of corporate structures and long-term strategy. That shift is strategically important: it helps build domestic industrial AI, reduces reliance on imported software, and strengthens the digital capabilities of the real economy amid intensifying global technological competition.
