Speaker
Description
Note: Please see attachment, it has figure and references
Data-driven methods are rapidly transforming experimental science [1], not only through advances in artificial intelligence but also by enabling scriptable access to scientific instruments [2]. Providing users with programmable control over instrumentation opens new opportunities for statistical analysis, adaptive experimentation, and more efficient materials discovery [3, 4, 5]. However, modern electron microscopy systems remain constrained by fragmented, vendor-specific hardware interfaces [4,5], limiting programmability and flexibility despite their substantial capital cost (often up to $10 million) [6].
In this talk, I will present a modular, vendor-agnostic framework for coordinating heterogeneous electron microscope hardware, where individual subsystems such as stage, scan, and detectors are abstracted as independent device-level services and orchestrated through a unified PyTango-based[6] control layer. This approach enables instruments to be accessed programmatically through scripting interfaces, allowing users to construct flexible, data-driven experimental workflows beyond traditional GUI-based operation. Figure 1 shows the schematic of the interface architecture and its realization in our university on a ThermoFisher spectra 300 microscope. Eventually our plan is to connect our interface with the BlueSky framework[8].
Code availability: https://github.com/pycroscopy/asyncroscopy
