Application of sensing techniques and artificial intelligence-based methods to laser welding real-time monitoring: A critical review of recent literature

Abstract Laser welding has been widely utilized in various industries. Effective real-time monitoring technologies are critical for improving welding efficiency and guaranteeing the quality of joint-products. In this paper, the research findings and progress in recent ten years for real-time monitoring of laser welding are critically reviewed. Firstly, different sensing techniques applied for welding quality monitoring are reviewed and discussed in detail. Then, the advanced technologies based on artificial intelligence are summarized which are exploited to realize varied objectives of monitoring such as process parameter optimization, weld seam tracking, weld defects classification, and process feedback control. Finally, the potential research problems and challenges based on real-time intelligent monitoring are discussed, such as intelligent multi-sensor signal acquisition platform, data depth fusion method and adaptive control technology. This fundamental work aims to review the research progress in laser welding monitoring and provide a basis for follow-on research.