Case Histories
Label control on steel blooms
Problem
Blooms are semi-finished steel products obtained during the hot rolling process. Once produced, they are labeled with an identification plate containing essential information (descriptive codes, QR Code, and Bar Code) for product traceability and quality control.
Starting point
The client is a major steel mill. The reading of identification tags during the picking phase is done manually. Due to the weight and large size of the blooms, the tag is often difficult to reach or even invisible before the picking process. Errors in selecting the wrong piece result in non-compliant final products or interruptions in the production process, leading to efficiency losses and additional costs.
To address this issue, a system was developed for the automatic identification of blooms and the reading of QR Codes printed on the metal identification plate applied to the cut section.
Solution Implemented
An end-to-end software application has been developed, based on AI-go and supported by a dedicated optical setup designed to automate the identification of blooms and the reading of QR Codes on the identification plates. The system consists of two fixed mirror-image optical units, strategically placed to capture images from different angles, even in low-light conditions and with high positioning variability.
By analyzing the images, the application can count and classify blooms based on their geometric profile, identifying the presence and position of the identification plates. The software then extracts the QR Code from each bloom’s tag, decoding it to obtain traceability and cataloging data, which is sent to the management system to verify that the selected piece matches the expected one.
Results
The implemented solution fully automates a previously manual and error-prone process, improving production chain efficiency, reducing picking times, and enhancing the accuracy of semi-finished product traceability.
The main results achieved are:
– automation of the bloom identification process, eliminating the need for manual intervention and reducing operational time;
– accurate QR Code reading, even when marks are present on the objects, improving semi-finished product traceability and reducing cataloging errors;
– elimination of the root cause of 75% of unexpected stoppages in the furnace loading roller path. Systematic identification of upstream process errors to enable global process improvement.
