Digitalized automated welding systems for weld quality predictions and reliability

Abstract The prevailing industrial and societal environment is driving welding manufacturing industries to employ green welding technologies for efficient, effective and reliable manufacturing and production. This paper presents current research work, whose aim is to develop a prototype of a commercial adaptive intelligent welding system with integrated weld quality attribute prediction and control. Supported by the study of scientific literature, initial results of experimental work which employed infrared thermography (IRT) based device and artificial intelligence (AI) system are discussed as a case study. Based on tested and validated welding samples, it is shown that the adaptive intelligent welding system being developed has self-monitoring capabilities for prediction of weld attributes, especially the depth of weld penetration, and has self-adjusting functionalities for weld control in off-line supervised conditions, and also can produce weldments of quality which conforms to EN ISO 5817. The findings imply that evolving welding technologies have practical industrial significance for monitoring and assurance, particularly as regards weld quality prediction and control, and, furthermore, as a tool to support decision-making when developing welding procedure specifications (WPS).