Development of AI-Assisted Quality Assurance for the Process of Ultrasonic Welding for High-Performance Thermoplastic Composites: A Review

Kurzfassung

Since ultrasonic welding of thermoplastic composites (TCs) gained attention from the industry in last years, the question arises, how artificial intelligence (AI) can help to establish the process in the industry and make it more reliable. This review summarizes the development of AI algorithms for the quality assurance (QA) of ultrasonically welded TCs. Focus is given to research dealing with spot welding before giving an overview about the newest developments in continuous ultrasonic welding (CUW): Many AI training approaches with different process parameters and diverse AI algorithms are described to make accurate quality predictions. One of the newer parameters for AI training is thermography. This parameter is described to be of huge potential for a reliable quality prediction in spot welding as well as in CUW and leads to AIs with high quality prediction quality. After an overview about the latest AI research, the review outlines the next steps which might be necessary in order to get a productive AI in the welding process of thermoplastic composites and describes the newest research for additional input data types.