Russian Metallurgy Moves Toward Intelligent Transformation
United Metallurgical Company (OMK) has unveiled its proprietary AI-powered scrap metal monitoring system at the AAAI 2026 international conference on artificial intelligence in Singapore.
OMK’s neural network-based solution automatically analyzes the type and contamination level of incoming raw materials, removing human subjectivity from the process. The technology has already been deployed at the casting and rolling complex of OMK’s Vyksa Steel Works, where scrap metal is delivered for steel production. Cameras continuously capture photos and video of incoming materials and feed the data into a system that analyzes every unloaded layer of scrap across all railcars and trucks.
At the core of the system are multiple computer vision models that process the entire material flow, isolate layers of scrap, assess contamination levels, and issue alerts that can automatically halt unloading. The system also determines whether incoming raw materials meet specified standards. In addition to detecting waste, the service can identify hazardous objects such as gas cylinders and barrels.
The innovative AI monitoring system enables faster and more efficient unloading of raw materials and shortens acceptance times, while eliminating errors caused by human judgment in scrap quality assessment. A key benefit is improved safety: inspectors now manage the process remotely, avoiding exposure to hazardous unloading zones. At the same time, the role of specialists is evolving, shifting toward deeper data analysis – evaluating system performance and fine-tuning AI algorithms.
International Recognition
The technology highlights the successful digitalization of production chains, a critical priority for the metallurgy sector, where measurable operational outcomes take precedence over marketing claims. The economic impact of the solution is estimated at hundreds of millions of rubles annually (several million US dollars per year). The innovation has also reduced the volume of waste received from scrap suppliers by three times and lowered CO2 emissions from electric furnace operations. This is particularly relevant as the industry moves toward greener metallurgy and circular economy practices.
According to the World Steel Association, scrap metal is essential to steelmaking and plays a key role in reducing CO2 emissions, conserving resources and energy, and supporting industry growth. Scrap accounts for up to 100% of raw materials in electric arc furnaces (EAF) and up to 30% in blast furnaces. In effect, every steel plant also operates as a recycling facility.
AAAI 2026 is one of the leading international scientific and applied conferences in artificial intelligence. Interest from participants at this global forum underscores the maturity of OMK’s solution and opens up export opportunities.
Smart Quality Control Across Production
Since 2021, several projects using machine vision for quality control have been under development and implementation at the Vyksa Steel Works, covering both production processes and finished products. For example, convolutional neural networks analyze video streams of hot-rolled coil production. This allows for more accurate detection of defects in steel strips and helps adjust casting parameters accordingly.
At the plant’s casting and rolling complex, surface quality assessment of hot slabs produced by continuous casting machines has been automated. Another application of machine vision at the facility is automated visual inspection to detect defects in couplings for casing pipes.
Billions in Savings and Growing Demand for Local Technologies
Across the sector, nearly all major Russian industrial companies are actively adopting Industry 4.0 strategies, moving from basic automation to advanced data-driven intelligence. For example, Norilsk Nickel is deploying AI across its operations and reporting significant economic impact – tens of billions of rubles annually. By 2030, the company aims to reach at least 50 billion rubles (approximately $670 million) in annual benefits.
Severstal is using more than 60 AI solutions to improve efficiency, generating over 1 billion rubles (approximately $13 million) in savings in 2024 alone. Magnitogorsk Iron and Steel Works (MMK) has established an AI competency center, delivering more than 4.5 billion rubles (approximately $60 million) in cumulative impact over five years. The sector is demonstrating measurable returns from digital transformation.
The deployment of OMK’s AI-powered scrap monitoring system is another example of strong demand for domestic technologies. The solution clearly demonstrates how artificial intelligence can address complex industrial challenges. Its successful rollout signals that the industrial sector is ready for intelligent digital transformation and opens new opportunities for developers of specialized software in metallurgy.
