How AI reduces laser joining costs

The text discusses the implementation of artificial intelligence (AI) in laser joining processes in a German factory of a U.S.-American car manufacturer. The factory wanted to eliminate a trim strip in the design of their new car model, which made the weld seam visible. To ensure high-quality visible seams, the project team decided to use laser brazing and partnered with Scansonic, a supplier of laser optics. Scansonic provided their ALO4 laser optics and a camera system for in-process observation of the weld seam. The camera system, equipped with AI, can identify defects in the weld seam, such as splatter or pores, and present them to the operator for repair. The implementation of the AI system required efficient data management and the development of a dashboard for data visualization. The AI system improved defect detection, allowing for the identification of smaller defects that are difficult to detect with the naked eye. This early detection saves time and money in production. The AI system also offers long-term quality control by storing videos of defects with car data for tracking and comparison. The digitization of processes and the use of AI in process control contribute to a fully digitized workflow and process stabilization. The implementation of the AI system took 18 months, but the learning curve suggests that future implementations will be faster. The AI system uses machine learning to analyze images of welded or brazed seams and identify patterns of defects such as spatters or pores. This analysis is called inference.