Functional Unit: A New Perspective on Materials Science Research Paradigms
This work addresses the problem of knowledge gaps in materials science for researchers transitioning to AI-driven methods, but it appears incremental as it builds on existing paradigms without presenting new experimental results or benchmarks.
The paper tackles the challenge of adapting materials science to data-driven AI approaches by introducing functional units (FUs) to bridge structure-property correlations and knowledge inheritance, aiming to integrate these concepts into a new AI-driven paradigm for materials design.
New materials have long marked the civilization level, serving as an impetus for technological progress and societal transformation. The classic structure-property correlations were key of materials science and engineering. However, the knowledge of materials faces significant challenges in adapting to exclusively data-driven approaches for new material discovery. This perspective introduces the concepts of functional units (FUs) to fill the gap in understanding of material structure-property correlations and knowledge inheritance as the "composition-microstructure" paradigm transitions to a data-driven AI paradigm transitions. Firstly, we provide a bird's-eye view of the research paradigm evolution from early "process-structure-properties-performance" to contemporary data-driven AI new trend. Next, we highlight recent advancements in the characterization of functional units across diverse material systems, emphasizing their critical role in multiscale material design. Finally, we discuss the integration of functional units into the new AI-driven paradigm of materials science, addressing both opportunities and challenges in computational materials innovation.