Next-Generation Friction Stir Additive Manufacturing: Emerging Technologies and Machine Learning for Process...

Friction Stir Additive Manufacturing (FSAM) has been recognized as a promising solid-state additive manufacturing (AM) technology for manufacturing metallic components with refined microstructures and enhanced mechanical properties. Unlike fusion-based AM processes, FSAM operates below the melting temperature, reducing solidification-related defects such as porosity, excessive residual stresses, and hot cracking. This review critically summarizes recent advances in FSAM, including tool design, thermal management, process parameter optimization, functionally graded structures, and multi-material fabrication. It also examines the integration of Artificial Intelligence (AI) and Machine Learning (ML) techniques, such as Artificial Neural Networks (ANN), Extreme Gradient Boosting (XGBoost), and Random Forest (RF), for process parameter prediction, property optimization, process monitoring, and defect detection. Current challenges associated with process scalability, data availability, tool wear, interlayer bonding, heat accumulation, and model generalization are critically discussed. Future research directions, including explainable AI, real-time closed-loop process control, digital twins, and standardized datasets, are also highlighted. This review provides an in-depth examination of the status of FSAM by integrating recent technological advances with intelligent manufacturing approaches and highlights key opportunities for sustainable, reliable, and data-driven solid-state additive manufacturing.