Automated Lab Speeds Up Material Development
Systematically investigating, evaluating, and specifically developing thousands of material variants—that’s what the new Energy Materials Acceleration Platform (E-MAP) at the Karlsruhe Institute of Technology (KIT) makes possible. It combines automated experiments with precise material characterization, thereby laying the foundation for accelerated development of energy materials.
Developing new functional materials often requires testing numerous material combinations. Conventional experiments quickly reach their limits when many variants need to be compared. The new Energy Materials Acceleration Platform (E-MAP) at KIT therefore automates key laboratory steps and combines them with modern artificial intelligence methods.
“Robotic systems handle tasks such as material preparation, sample handling, thin-film deposition, and characterization,” says Dr. Holger Röhm of KIT’s Light Technology Institute (LTI), whose research team developed the platform. “This allows experiments to be conducted with high precision and reproducibility. With E-MAP, we can more quickly identify which material compositions and fabrication conditions are particularly promising for a specific application.”
Modular Design and Flexible Expandability
The E-MAP is designed as a closed system and enables the automated fabrication and analysis of materials under controlled conditions. Thanks to its modular design, the platform can be continuously expanded to include new experiments and analytical methods. “A key advantage of the E-MAP is its modular design,” says Professor Alexander Colsmann of KIT’s LTI. “We can integrate new experiments and characterization methods, thereby adapting the platform to different scientific questions. Collaboration partners from academia and industry can also contribute their own methods and equipment.”
From Automated Experiments to AI-Driven Materials Development
Automation generates large amounts of experimental data. To use this data to develop new materials more quickly and in a more targeted manner, scientists led by Professor Pascal Friederich are researching AI methods that enable the platform to learn from prior knowledge and independently suggest promising next experimental steps.
“With modern material systems, the search space is often so large that not every variant can be investigated experimentally. That’s why we’re developing methods that use artificial intelligence to learn from previous results and specifically select the most informative next experiments. This allows us to reduce the number of necessary experiments and significantly shorten the path to new materials,” says Pascal Friederich, head of the Artificial Intelligence for Materials Science group at the KIT Department of Informatics.
The AI combines data from the automated lab with scientific prior knowledge and insights from the scientific literature. On this basis, it can identify particularly promising material combinations and strategically guide new experiments. This allows large material spaces to be explored much more efficiently and promising candidates to be developed more quickly for practical applications.
About E-MAP
The Energy Materials Acceleration Platform (E-MAP) is a Self-Driving Lab (SDL), i.e., a largely autonomous research platform. Such SDLs are a central component of the German High-Tech Agenda. Construction of E-MAP began in 2023. To date, KIT has provided approximately 600,000 euros for the platform’s technical equipment. The Carl Zeiss Foundation has supported the development in recent years as part of the KeraSolar research project. E-MAP is integrated into the planned Helmholtz Acceleration Alliance (HELMA), which aims to network autonomous research platforms in the materials and life sciences across multiple Helmholtz Centers.