Application of artificial intelligence in additive manufacturing (3D printing)

Abstract

The paper presents a comprehensive analysis of artificial intelligence application to improve additive manufacturing quality. Systemic problems of modern 3D printing are identified: quality instability, high defect probability, and labor-intensive model preparation. Six key AI-based quality control methods are analyzed: digital twin, intelligent monitoring, predictive modeling, defect classification, predictive maintenance, and generative design. The combination of deep learning methods (CNN, U-Net, LSTM/GRU) with classical machine learning algorithms (XGBoost, Random Forest) and explainable AI technologies (XAI) ensures high defect detection accuracy and decision interpretability. Particular attention is given to the digital twin concept as an integrating platform for creating a dynamic virtual replica of the printing process. The results confirm that AI integration represents a fundamental transformation of approaches to design, production, and quality control in additive technologies.