A crane is automatically loading wire-rod products onto a trailer for shipment. /Courtesy of POSCO DX

POSCO Group said on the 1st that it has developed technology that combines artificial intelligence (AI), information technology (IT), and automation to automatically transport nonstandard products with cranes and will apply it at steelworks sites.

Using the newly developed crane automation technology, wire rod coils weighing up to 8 tons at a time can be transported accurately and safely. Wire rod coils, which are thinly drawn steel wound into a spiral shape, are nonstandard products with varying sizes and packaging forms, making shape recognition difficult. As a result, when moved by crane, they sway significantly, and the area has been regarded as difficult to automate.

POSCO DX said it solved this technical challenge by integrating vision AI and precise position control technology in cooperation with POSCO.

To move shipped wire rod coils according to work orders issued from the manufacturing execution system (MES) at the steelworks, a LiDAR sensor attached to the crane is used to locate the coil. Then, AI-based vision technology finds the virtual center point of the wire rod coil and recognizes its outer diameter, inner diameter, and width. Depending on the number of coils, the hook entry position and depth are automatically calculated and linked to control. Field verification showed the technology achieved ±20 mm accuracy and a 99.5% detection rate.

POSCO DX said, "By precisely measuring the shape of wire rod coils even when the loaded coils are toppled or pressed by other products, stable work automation is possible." POSCO DX also developed a safety management system to prevent accidents in which workers or vehicles collide with cranes in the wire rod warehouse, strengthening on-site safety levels. This method attaches LiDAR sensors to major traffic sections of the product yard to accurately recognize people and vehicles and automatically control crane operation when there is a collision risk with the crane. The project applied a 3D LiDAR sensor from the LiDAR specialist "Aiyobot."

POSCO plans to use this successful experience of building crane automation technology to spread the technology to other sites within the steelworks. It is also pushing on-site verification of technologies such as sheet-count control automation, developed to implement automation of plate cranes, another type of nonstandard product.

A POSCO official said, "We have successfully developed crane automation technology for nonstandard products, which had been considered difficult to automate," adding, "We will actively work to solve challenges in manufacturing and logistics sites through the advancement of digital transformation (DX) technology using AI."

※ This article has been translated by AI. Share your feedback here.