RUSAL Confirms the Maturity of Russian Computer Vision Solutions
Russian aluminum producer RUSAL plans to complete deployment this year of a machine-vision-based electrolysis monitoring system across five of its smelters.
The solution was developed in-house and marks the holding’s first industrial use of artificial intelligence in production. It targets environmental performance and stands as the company’s most capital-intensive AI initiative to date. Total rollout costs are estimated at about 1.6 billion rubles (about $21 million). The project enables rapid detection of anomalies, reduces environmental impact, and sets a new benchmark for digitalization in the sector.
The computer vision system installs dedicated cameras to monitor roughly 6,500 electrolyzers around the clock. AI models detect seal integrity breaches in real time and automatically alert operators. In practice, the new monitoring system cuts the duration of depressurization events by half compared with manual checks performed during scheduled walk-throughs in electrolysis shops.
The technology is already fully deployed in electrolysis buildings at the Krasnoyarsk Aluminum Smelter (KrAZ). Equipment installation continues at the Bratsk, Volgograd, Irkutsk, and Novokuznetsk smelters.
Neural Networks Safeguard Environmental Performance
RUSAL is a leading global aluminum producer and one of the largest suppliers of low-carbon aluminum. The company operates an integrated value chain, from bauxite mining to higher-value finished products, which provides tight quality control across all stages. In 2025, RUSAL sold 4.49 million metric tons of aluminum and alloys, up 16% year over year from 2024.
More than 90% of RUSAL’s output uses renewable power, while ongoing innovation reduces emissions at every stage. That approach enabled the company to launch one of the world’s earliest “green” aluminum lines under the ALLOW brand, combining process technology with environmental performance.
Import-Independent Digitalization
The pilot for machine-vision-based electrolysis monitoring was developed between 2018 and 2020 and became the company’s first industrial AI application in aluminum production. In April 2026, the Engineering and Technology Center RUSAL (RUSAL ITTs), which developed the technology, won the “Best IT Solutions for the Metallurgical Industry” award in the ESG – Environment and Industrial Safety category.
RUSAL ITTs was established in 2002 in Krasnoyarsk and serves as the company’s primary R&D hub. The center conducts research on production technologies and waste processing, develops new equipment, and advances AI tools across multiple stages of aluminum manufacturing.
For Russia’s IT sector, RUSAL’s use of computer vision to monitor electrolysis demonstrates that locally developed AI solutions can be integrated into heavy industry at scale. More broadly, the project signals the maturity of Russia’s push toward industrial AI and import-independent technologies.
From Alumina to Electrolysis
At the same time, RUSAL continues to deploy AI to optimize other process stages. In 2024, the Sayanogorsk Aluminum Smelter tested a neural-network-based electrolysis control system that determines the chemical composition inside the electrolyzer and adjusts the smelting process in real time. The system accelerated production and reduced manufacturing costs.
In 2025, the company completed trials of a machine-vision system for melt preparation control. Cameras and neural networks track operator actions and equipment performance, displaying step-by-step prompts based on process maps (KPVO). Depending on alloy complexity and ingot type, preparation involves between 7 and 20 operations, each performed at a precise time and under strict process parameters. AI reduces human error, improves control over alloy chemistry, and supports efficient slag removal. The company plans to scale the technology across all RUSAL plants, improving metal quality and production stability.
In 2026, RUSAL deployed an in-house neural-network system to control aluminum hydroxide particle fractions. The approach improves alumina quality and the aluminum produced from it, while reducing energy consumption during electrolysis. Earlier, RUSAL was among the first global producers to introduce digital twins at its alumina refineries to optimize processes, lowering resource use and increasing productivity.
From Point Solutions to an Ecosystem
RUSAL’s portfolio reflects a broader trajectory in Russian metallurgy, where companies are moving from point computer vision tools toward intelligent control of production processes. That shift is set to become a key driver of technological independence and global competitiveness in the industry.
