Researchers, technology developers and representatives from the European steel industry gathered in Freiburg, Germany, last week for the AID4GREENEST workshop “Materials Characterisation for Industrial R&D: From Fragmented Data Towards AI-powered Workflows.”
Hosted by Fraunhofer IWM and co-organised by AID4GREENEST project partners ePotentia, Ghent University and Fraunhofer, along with the European Materials Modelling Council (EMMC), the two-day event showcased how digitalisation, artificial intelligence and FAIR data principles are transforming materials characterisation for greener steel production.
As advanced characterisation techniques generate ever-increasing volumes of experimental data, one of the greatest challenges facing industry is ensuring that this information is organised, accessible and reusable.
The workshop focused on practical solutions for overcoming fragmented data workflows by introducing standardised approaches that enable more efficient research, improved traceability and AI-driven analysis.
The first day explored the foundations of digital materials data management, highlighting the importance of FAIR (Findable, Accessible, Interoperable and Reusable) principles and the CHADA (Characterisation Data) framework for standardising experimental workflows. Participants learned how structured data can support machine learning models capable of predicting long-term material behaviour, reducing testing times and accelerating alloy development for industrial applications.
A guided tour of the Fraunhofer IWM laboratories provided attendees with the opportunity to see the complete digital workflow in practice, from mechanical testing and microstructural characterisation through to the creation of AI-ready datasets.
Interactive afternoon sessions enabled participants to work directly with standardised test data, apply digital documentation frameworks and explore how Large Language Models (LLMs) can be used to query materials databases using natural language.
The day concluded with a discussion led by the EMMC’s Dr. Alexandra Simperler on the opportunities and challenges surrounding the adoption of digital materials infrastructures across industry.
The second day shifted the focus towards practical AI tools for microstructural analysis. Participants were introduced to ePotentia’s MicrostructureDB platform, an AI-enhanced environment designed to organise, analyse and share microstructural datasets in a standardised and interoperable format. Through a series of guided demonstrations, attendees explored how artificial intelligence can assist with organising microscopy data, automatically enriching metadata and accelerating the interpretation of complex materials datasets.
Hands-on sessions also demonstrated the growing role of conversational AI in materials science, allowing users to query databases using natural language and rapidly identify relevant microstructural information.
Additional workshops showcased explainable AI methods for analysing optical microscopy, scanning electron microscopy (SEM) and electron backscatter diffraction (EBSD) data, highlighting how machine learning can support faster, more reproducible materials characterisation.
Throughout the workshop, discussions emphasised that digitalisation is not simply about managing larger volumes of data, but about creating connected workflows that enable faster decision-making, stronger collaboration between research and industry, and ultimately more sustainable steel technologies.