Accelerating Ceramic Materials Development via Machine Learning: A Critical Review

Abstract

Ceramics are inorganic, non-metallic materials renowned for their exceptional thermal, mechanical, chemical, and functional properties, making them indispensable across aerospace, biomedical, electronic, and energy applications. Despite their advantages, optimizing ceramic processing remains a major challenge due to the complex interplay of temperature, particle characteristics, atmosphere, and additives, where minor deviations can significantly affect microstructure and performance. Traditional experimental and computational approaches are often time-consuming or resource intensive. Machine learning (ML) offers a transformative alternative by enabling accurate property prediction, process optimization, materials discovery, and degradation analysis. This review presents a comprehensive overview of ML applications in ceramics, literature examples, and discusses current challenges and future directions, highlighting the potential of ML to accelerate the design and development of high-performance ceramic materials