Artificial intelligence for scanning probe microscopy: From data analysis to autonomous experimentation

Abstract

Abstract Advanced scanning probe microscopy (SPM) measurements have traditionally required highly specialized skills, but the integration of artificial intelligence (AI) and machine learning has begun to transform how SPM is operated. This review surveys AI methods applied to SPM, organized by technique rather than chronology: unsupervised learning for feature extraction, support vector machines for real-time classification, convolutional neural networks for image analysis, Bayesian optimization for experiment planning, reinforcement learning for atomic and molecular manipulation, and object detection for feature localization. The emerging use of large language models as natural-language interfaces for microscope operation is discussed as a distinct topic. For each method, the operating principles and representative SPM applications are presented. This method-centric organization enables direct comparison of AI approaches applied to similar SPM tasks and clarifies the complementary roles---perception, decision-making, control, and orchestration---that each technique plays in integrated autonomous SPM systems.