POSCO Gwangyang Steelworks Achieves "AI-Driven Blast Furnace Automation" Using Proprietary Technology

Achieving Both Safety and Quality with AI-Powered Dispatch Automation System

Aiming to Realize a Smart Factory in Steelmaking by Advancing Dispatch Automation and Expanding Application Across All Plants

[Asia Economy Honam Reporting Headquarters, Reporter Heo Seon-sik] The Steelmaking Department of POSCO Gwangyang Steelworks (Head Kim Ji-yong) has developed an automated tapping system using artificial intelligence (AI), marking an innovative start for the smart factory in the steelmaking sector.

In particular, this automated tapping system has attracted more attention because it achieved tapping automation using POSCO's proprietary technology from the beginning without relying on external technologies.

Tapping is the process of transferring molten iron, which has undergone impurity removal, into a transfer container. During tapping, there were quality inconsistencies due to variations among operators in the timing of ferroalloy input for oxygen and component adjustment of the molten iron, as well as the tapping end time.

To resolve this, the Steelmaking Department at Gwangyang Steelworks collaborated with POSCO Technical Research Institute, Electrical Technology Section, POSCO ICT, and POSTECH Graduate School of Steel/Energy Materials. First, the work was subdivided step-by-step, and data under various conditions were collected and standardized. Then, a standard tapping pattern was derived, and based on this, tapping automation was implemented.

Based on this, Gwangyang Steelworks achieved not only improved work efficiency and quality stabilization but also significantly enhanced worker safety.

In the 2nd Steelmaking Plant where the tapping automation is applied, tapping equipment can be automatically controlled with the push of a button. Additionally, based on the derived tapping pattern, the timing for Dart input to prevent ferroalloy and slag outflow is determined, and the entire tapping process is automatically completed, reducing discrepancies among operators.

Moreover, through AI video recognition technology, it is now possible to detect risks from high-temperature operations without visually monitoring the tapping process directly. AI analyzes the video to automatically detect hazards, and especially for the risk of ladle overflow, risk detection and control technology using ladle camera video analysis is applied.

The Steelmaking Department at Gwangyang Steelworks plans to further advance the tapping automation algorithm and develop AI video recognition technology to implement automatic control for all potential risks during the tapping process. Furthermore, they intend to expand this system to all converters in the steelmaking plants to significantly reduce the risk of worker safety accidents.

Lee Seong-jin, Manager of the 2nd Steelmaking Plant who led the system development, said, “I am truly grateful to the POSCO Technical Research Institute, Electrical Technology Section, POSCO ICT, and POSTECH Graduate School of Steel/Energy Materials who worked together with the determination to achieve tapping automation with our own technology. With the pride of accomplishing an important milestone in steelmaking history as our driving force, we will realize tapping automation across all steelmaking plants to ensure our colleagues can work more safely and comfortably.”

Meanwhile, as a global steel company integrating the 4th Industrial Revolution, POSCO actively supports the digital competency development of its employees through various in-house training programs ranging from basic IT knowledge to expert courses in artificial intelligence and big data utilization.

This content was produced with the assistance of AI translation services.