Integration of Artificial Intelligence for Real-Time Monitoring and Process Control in Metal Additive Manufacturing Systems

Abstract

Metal additive manufacturing (AM) has attracted significant attention because of its ability to produce complex geometries, reduce material waste, and provide greater design flexibility than conventional manufacturing methods. However, its wider industrial use is still limited by process instability and defect formation during fabrication. Defects such as porosity, lack of fusion, residual stress, and distortion can reduce part quality and reliability, making real-time monitoring and process control increasingly important. This paper reviews recent progress through 2024 in the use of artificial intelligence (AI) for real-time monitoring and control in metal AM systems. The review shows that AI has improved anomaly detection, melt-pool analysis, defect prediction, and adaptive control performance. It also highlights the growing role of sensor fusion and low-latency computing in supporting in-situ decision-making. Despite these advances, challenges remain in data availability, model generalization, interpretability, and industrial reliability. Overall, AI is playing an important role in advancing metal AM toward more intelligent, stable, and quality-assured manufacturing.