EXPLAINABLE ARTIFICIAL INTELLIGENCE FOR QUALITY MONITORING IN METAL 3D PRINTING

Abstract

Metal 3D Printing has become a key manufacturing technology for producing lightweight, geometrically complex, and customized components across aerospace, automotive, biomedical, and industrial sectors. However, manufacturing defects such as porosity, cracking, residual stress, lack of fusion, thermal distortion, and surface irregularities continue to affect component quality and production reliability. Conventional Artificial Intelligence-based quality monitoring systems often operate as black-box models, making it difficult for engineers to understand the reasoning behind quality predictions and corrective recommendations. Recent advances in Explainable Artificial Intelligence (XAI), Machine Learning, Industrial Internet of Things (IIoT), Digital Twin technology, and predictive analytics enable transparent, interpretable, and trustworthy manufacturing quality monitoring. This paper proposes an Explainable Artificial Intelligence framework for quality monitoring in Metal 3D Printing by integrating interpretable Machine Learning models, realtime sensor analytics, cloud manufacturing, Digital Twin environments, and intelligent visualization. The proposed framework continuously predicts manufacturing quality while providing explainable insights into process behavior and defect formation. Experimental analysis demonstrates significant improvements in defect prediction transparency, manufacturing quality, process stability, operational efficiency, and decision reliability compared with conventional blackbox quality monitoring approaches. Keywords: Explainable Artificial Intelligence, Metal 3D Printing, Quality Monitoring, Machine Learning, Smart Manufacturing, Industrial Internet of Things, Digital Twin, Predictive Analytics, Industry 4.0, Intelligent Manufacturing.