The Metinvest Group is employing artificial intelligence at its enterprises to improve workplace safety and oversee product quality. This was shared by Maksym Balaniuk, head of the R&D centre at Metinvest Digital, during the "Forbes Tech 2024: Technology for Your Business" conference held on 22 November in Kyiv.
At this major tech event of the year, organised by "Forbes" magazine, participants explored ideas on leveraging technology to achieve business objectives under resource constraints while staying aligned with global technological trends. The conference featured four thematic sessions with 57 speakers, including government officials and representatives from companies such as Meta, Microsoft, and Google.
Maksym Balaniuk participated in the discussion panel titled "ChatGPT – Two Years On: How Ukrainian Companies Have Evolved Their Use Cases Over This Time?" Other contributors included representatives from Darnytsia, KNESS, and Favbet Tech.
According to Balaniuk, Metinvest Group employs a computer vision-based solution in its occupational safety efforts. The platform, known as SPAIS, integrates with industrial video surveillance systems. "Different models operate on various cameras, collecting data on safety violations or compliance. In this way, we monitor whether individuals are in safe or hazardous areas and whether they are equipped with personal protective equipment. We are continuously adding new functionalities. Soon, the system will be able to detect and alert on the presence of fire, smoke, or smoking in hazardous areas," he explained.
Balaniuk added that this platform is already being effectively utilised at several Metinvest enterprises and is planned to be scaled to other company assets. "The impact is clear: the number of violations has decreased," he noted.
Another new product from Metinvest Digital is a computer vision-based solution designed to analyse and monitor product quality, reducing material waste and ensuring maximum compliance with customer requirements. This, in turn, helps lower costs associated with complaints and reprocessing of metals.
"The currently available first version of the product is functioning at the level of a highly skilled expert. Next year, we plan to release an enhanced version that will be even more efficient, surpassing human capabilities and providing superior, fatigue-free quality control," Balaniuk added.
He stated that the first version of the solution is expected to deliver an economic benefit of approximately $250,000 annually for a single enterprise. The next version is projected to double these savings.
