The Institute of Energy Technologies - Fundamental Electrochemistry (IET-1) focuses on the development of performance oriented and sustainable materials and components for the electrochemical energy storage and conversion. Aiming to develop sustainable and cost-effective batteries, fuel cells and electrolyzers with improved energy and power density, longer lifetime at maximal safety is the challenge of the projects. These key technologies drive forward the energy transition and structural change in the Rhineland region. Further info on our exciting projects: https://go.fzj.de/IET-1

YOUR JOB:

The Innovationpool Project “Data for Technology Assessment” (DaTA) aims to create a comprehensive, publicly accessible repository for technology data to support research in energy systems. This project focuses on advancing the TechDB database with AI-driven automated data collection, developing methods and tools for integrating heterogeneous data into multi-energy system design and operation, and creating reference test cases for comparative evaluation of new methods and algorithms.

We are seeking a Research Associate specializing in Machine Learning to contribute to the digital analysis and reconstruction of Solid Oxide Cell (SOC) electrode microstructures. This role is part of the Electrochemical Processing and System Technology department at IET-1, where our team is working to automate electrode analysis, specifically Focused Ion Beam-Scanning Electron Microscope (FIB-SEM) imaging, through data assimilation and model calibration. The goal is to develop physics-informed neural network models for electrodes and integrate these models as machine learning-based surrogate models for stack and system-level optimization. Your tasks in detail:

• Develop automated data processing pipelines to analyze SOC electrodes using physics-informed neural networks and related approaches
• Collect and process existing FIB-SEM images to digitally reconstruct and regenerate electrode microstructures using Generative Adversarial Networks (GANs) or similar techniques. Image segmentation may be required.
• Train and validate ML-based surrogate models using both experimental data (e.g., Electrochemical Impedance Spectroscopy - EIS) and numerical simulations
• Collaborate with numerical and experimental teams to define requirements and provide technical support
• Perform numerical simulations to support SOC stack and system design optimization
• Document and publish research findings in scientific journals and present at conferences

• Develop automated data processing pipelines to analyze SOC electrodes using physics-informed neural networks and related approaches
• Collect and process existing FIB-SEM images to digitally reconstruct and regenerate electrode microstructures using Generative Adversarial Networks (GANs) or similar techniques. Image segmentation may be required.
• Train and validate ML-based surrogate models using both experimental data (e.g., Electrochemical Impedance Spectroscopy - EIS) and numerical simulations
• Collaborate with numerical and experimental teams to define requirements and provide technical support
• Perform numerical simulations to support SOC stack and system design optimization
• Document and publish research findings in scientific journals and present at conference