Integration of scanning probe microscope with high-performance computing: Fixed-policy and reward-driven workflows...

Area of Science:

• Materials Science
• Nanotechnology
• Computational Science

Background:

• Scanning Probe Microscopy (SPM) is crucial for nanoscale characterization.
• Automating SPM operations can accelerate scientific discovery.
• Integrating machine learning (ML) with SPM requires robust computational infrastructure.

Purpose of the Study:

• To create a Python interface for controlling SPM systems.
• To enable the use of high-performance computing for ML-driven SPM workflows.
• To establish a platform for automating scientific discovery using SPM.

Main Methods:

• Development of a Python interface library for SPM control.
• Integration with local and remote high-performance computing resources.
• Implementation of a platform for abstracting SPM operations into ML workflows.

Main Results:

• A functional Python interface enabling seamless SPM control.
• Demonstrated capability to leverage high computation power for ML algorithms.
• A versatile platform supporting automated SPM workflows.

Conclusions:

• The developed infrastructure facilitates automated SPM operations.
• This system enables autonomous scientific discovery through machine learning.
• The platform supports both routine SPM tasks and advanced research.