Machine learning-aided material selection, parameter optimization, and process monitoring in laser powder bed fusion and laser directed energy deposition with powder feed

Laser-based metal additive manufacturing (AM) technologies, such as laser powder bed fusion and laser-based directed energy deposition, have been broadly adopted in a wide range of industries, including the aerospace, automotive, biomedical, and energy sectors. These technologies require cross-scale coordination spanning six orders of magnitude, involving complex interactions between laser parameters, feedstock properties, and thermal-fluid dynamics. The resulting high-dimensional datasets span materials, processing parameters, laser-metal interaction, and final part properties, posing significant challenges to understanding intricate processing-structure-property (P-S-P) relationships. Recent advances in machine learning (ML), with its capacity to uncover hidden patterns and correlations from large, high-dimensional datasets, present a transformative approach to accelerate discovery and optimization in metal AM. In this review, we provide a comprehensive overview of recent advances and challenges in the application of various ML methods in two major laser-based metal AM processes: laser powder bed fusion for metal and laser directed energy deposition with powder feed. We examine the application of ML in three main aspects: (1) material selection, including AM-tailored materials development and metal powder selection; (2) optimization of processing parameters through the development of process maps and optimization strategies; and (3) in situ monitoring and quality assurance in AM processes using sensor-based systems, including acoustic emission, visible light, infrared, synchrotron X-ray, and multi-sensor approaches. Moreover, we examine the characteristics, limitations, and data requirements of various ML models in each of these domains, guiding model selection and experimental design. By adopting an ML-centric perspective, this review aims to inform future research directions, support the development of more autonomous metal AM systems, and accelerate the optimization of laser-based metal AM.