During production of cold-rolled steel strips, the arc welding machine plays a pivotal role in maintaining seamless operation of the production line. It is responsible for welding the ends of steel strips together, ensuring uninterrupted flow throughout the process. Its operational stability is critical, as any functional failure would result in severe disruptions and heavy financial damages.​

However, complexity of the production environment, coupled with unpredictability of welding machine and equipment, often caused  inconsistent welding quality. The absence of effective online quality inspection system further contributes to occurrence of defects and damage incidents.​

Currently, workers rely on traditional methods such as referencing machine’s past data, past experience, manual impact tests and visual inspections to ensure weld quality; of which, such methods are highly inaccurate and not in real-time, leading to inefficiency and increased risk of quality defects, from time to time. 

•   Visibility of the welding process, combined with big data welding parameters to give weld seam instructions. ​
•   System analysis for weld seam quality after welding completion​
•   Optimal welding parameters given by AI model​.
•   AI-powered intelligent welding system (infra-red weld seam detector, big data welding recommendation model system, welding big data closed-loop system) achieves fully automatic closed-loop re-welding of defective weld seams​.
•   The welding model iteration does not require manual intervention, and can perform online self-learning and self-adaptation​