Navigating the Microscopic World—Synergy of Human Expertise, ML Efficiency and API Enabled Automation

Microscopy has become a standard tool in characterization laboratories and greatly enhanced our understanding of nanoscale structure-function relationships. However, traditional microscopy operations often rely on manual, human-centric approaches. To overcome these limitations, we present the integration of application program interface (API) with machine learning. We developed AEcroscoPy, a cross-platform Python API designed to automate microscopy experiments. The integration of API-driven automation, human expertise, and machine learning (ML) efficiency showcases the power for accelerating scientific discovery. Our development of ML driven automated and autonomous experiment (AE) in scanning probe microscopy (SPM) facilitates the exploration of material functionalities and mechanisms. Using AE-SPM, we discovered coexistence and interplay of two ferroelectric subsystems in wurtzite ferroelectric thin films. By employing ML-driven approaches, we have investigated phenomena such as domain wall dynamics and switching mechanisms in ferroelectric materials, as well as the interactions between domain structures and local properties. By incorporating physical hypotheses in active learning model, our approach has enabled the microscope to autonomously discover the physical laws influencing domain switching. This approach not only enriches our understanding of material properties at the nanoscale but also opens new avenues for the application of ML in experimentation. Although these methodologies were applied to specific materials, they possess broad potential to revolutionize various characterization techniques.