An Automated Scanning Transmission Electron Microscope Guided by Sparse Data Analytics

TL;DR

This paper presents a novel AI-guided automated STEM platform that uses sparse data analytics and machine learning for real-time decision-making, enabling high-throughput, automated material analysis.

Contribution

It introduces a closed-loop control system integrating sparse data analytics and machine learning to automate STEM experiments and feature detection.

Findings

Demonstrated real-time automated decision-making in STEM

Enabled high-throughput material feature analysis

Improved interpretability of feature detection

Abstract

Artificial intelligence (AI) promises to reshape scientific inquiry and enable breakthrough discoveries in areas such as energy storage, quantum computing, and biomedicine. Scanning transmission electron microscopy (STEM), a cornerstone of the study of chemical and materials systems, stands to benefit greatly from AI-driven automation. However, present barriers to low-level instrument control, as well as generalizable and interpretable feature detection, make truly automated microscopy impractical. Here, we discuss the design of a closed-loop instrument control platform guided by emerging sparse data analytics. We demonstrate how a centralized controller, informed by machine learning combining limited $a$ $priori$ knowledge and task-based discrimination, can drive on-the-fly experimental decision-making. This platform unlocks practical, automated analysis of a variety of material features, enabling new high-throughput and statistical studies.