Invention Reference Number
Atomic force microscopy (AFM) is a powerful tool for nanoscale characterization, but it is limited by slow scanning speeds, small imaging areas, and the need for expert operation both during image acquisition and post-processing. SimuScan makes AFM faster, smarter, and more autonomous by generating realistic, labeled synthetic data that enables and accelerates the training of artificial intelligence models capable of recognizing nanoscale features without the need for tedious and time-consuming manual labeling. This innovation enables high-throughput imaging and adaptive scanning with minimal human supervision, expanding AFM’s reach into manufacturing, diagnostics, and materials discovery.
Description
SimuScan is a synthetic data generation and control platform that transforms atomic force microscopy (AFM) into an autonomous and scalable technology. It produces realistic, AFM-like datasets of diverse nanostructures and biological specimens, embedding lifelike imaging artifacts, typical of those observed in AFM, and paired labels for machine learning applications. The resulting data can train AI models that recognize, segment, and classify features in real experimental images without expert annotation. These models can be used in real time for decision-making and adaptive exploration, or for the post-analysis of large datasets required for wide-area nanoscale mapping. Once trained, these models integrate into an adaptive AFM workflow capable of identifying features in real time, performing zoom-ins, and scanning large areas automatically. This allows AFM systems to operate with minimal human input while maintaining nanoscale precision. By removing the bottleneck created by the scarcity of manually labeled datasets and combining AI with intelligent scanning control, SimuScan drives AFM toward the next generation of autonomous and intelligent systems—capable of exploring large areas, locating and characterizing different targets at high resolution, and strengthening AFM’s position in both research and industry.
Benefits
- Reduces dependence on expert users and manual data labeling.
- Enables real-time adaptive scanning and fully autonomous control.
- Expands imaging coverage, allowing statistically significant surveys.
- Improves accessibility, efficiency, and throughput for nanoscale studies.
Applications and Industries
- Materials science and nanotechnology research
- Semiconductor and electronics manufacturing
- Biomedical diagnostics and pathogen detection
- Energy and advanced materials development
Contact
To learn more about this technology, email partnerships@ornl.gov or call 865-574-1051.
