Treffer: Data integration and data fusion approaches in self-driving labs: A perspective.

Self-driving laboratories (SDLs) are transforming materials discovery by combining automation, machine learning, and real-time feedback. Yet, their success depends on robust data integration and fusion methods capable of handling materials data that are heterogeneous, sparse, and multi-scale. Such data span theoretical models, simulations, and experimental techniques across diverse spatial and temporal scales, creating significant challenges for interoperability and analysis. This perspective reviews the state-of-the-art techniques, including knowledge graphs, structured pipelines, multimodal machine learning, and physics-informed models, that are enabling materials science and SDLs to unify and learn from disparate data sources, identify critical challenges, and propose forward-looking directions to enhance data readiness, interoperability, and predictive power in SDLs. We also highlight emerging methods such as transformer architectures, zero-shot learning, and real-time stream processing, and discuss the critical need for more scalable, interpretable, and adaptive solutions to fully realize autonomous materials innovation. By mapping out both the current landscape and future opportunities, we argue that next-generation data integration and fusion are not just enablers but essential pillars for achieving fully autonomous, adaptive, and intelligent SDL systems capable of addressing the complexities of hierarchical and multifunctional materials. [ABSTRACT FROM AUTHOR]

Copyright of APL Machine Learning is the property of American Institute of Physics and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)