easyTEM: Developing Resource-Efficient Transmission Electron Microscopy for Democratizing Problem Solving in Materials Science

easyTEM develops a software platform that improves the usability and utilization of transmission electron microscopes (TEM) and increases resource efficiency. Idle instrument time is used to learn and optimize alignment, configuration, and acquisition workflows automatically. The project combines deep reinforcement learning for parameter and process optimization with LLM-based, user-centered assistance, for example natural-language interaction, adaptive interfaces, and safety safeguards. A further focus is vendor-agnostic control, including GUI-parsing and GUI-control agents when APIs are unavailable or costly, as well as safe remote operation for novice and occasional users. Key components and models will be released as open-source software.