Pipe Metallurgical Company (TMK) presented its own developments in the field of artificial intelligence, used to increase the efficiency of electric steel production, at the TAdviser SummIT industry event. The solutions presented by TMK are designed to account for and control the quality of scrap metal - the main raw material for the production of pipe blanks.
As the head of the machine vision department at the TMK Center for Digital Technologies Alexander Dunaev said, the company chose its own solutions after analyzing existing ones on the product market. According to him, ready-made technologies need to be adapted for specific production, and after successful implementation at one enterprise, they require modification for use at another. The scaling potential of proprietary software products, originally designed to solve TMK problems, is much higher.
The key tool for fixing and monitoring scrap metal is machine vision technology. First of all, TMK uses it to estimate the volume of available raw materials. “We use a system of cameras, the data from which is processed using special software, which allows us to measure the geometry of the surface and draw maps of each of the sites with scrap metal. This is necessary, among other things, to optimize the accounting process, prepare work on moving scrap and train personnel,” said Alexander Dunaev.
Another component of the complex system is tracking products entering the electric furnace melting shops. Video cameras record the entry of vehicles with scrap into production premises and allow you to track the movement of raw materials through processing stages. This information is loaded into software, thanks to which company employees have access to a complete history of the receipt and use of scrap metal.
To control the quality of raw materials, TMK also uses artificial intelligence. “It is important for us that there are no inclusions in the scrap that could disrupt the technological process. To search for such items, we selected an algorithm based on convolutional neural networks, which we trained by analyzing photographs. Now we can identify unnecessary objects in scrap and remove them from the total mass so that they do not end up in the smelting,” noted Alexander Dunaev.
According to him, the introduction of digital technologies not only improves economic indicators, but also helps reduce emissions and improves environmental friendliness of production.
Artificial intelligence is used at TMK to solve a wide range of problems. In electric steelmaking, in addition to accounting and control of raw materials, it helps to optimize the selection of materials and melting parameters; in pipe rolling shops, it monitors the condition of equipment and tools, the technological chain, parameters and quality of finished products, and is used when sorting and sending orders. AI is also used in office processes - from commerce to accounting.
TMK presented its developments in the field of AI for electric steel production
