Finding efficient ways to test digital materials in industry, contract laboratories, and research institutions is becoming more urgent in light of the digital transformation in business and science. “The exponentially increasing use of artificial intelligence and the machine learning involved in it in all sectors of the German economy has also found its way into materials science and materials engineering,” explains Prof. Martina Zimmermann, who heads the Materials Characterization and Testing competence field at Fraunhofer IWS in Dresden. “We can use digitally recorded and structured materials data to achieve competitive advantages in everyday laboratory work and industrial companies.”
 

Popular Digital Lab Books

As a result, more and more materials science, biotechnology, and pharmaceutical companies and institutes are replacing their traditional paper lab notebooks, in which preparation steps, experimental results, and processes for producing materials and active ingredients used to be recorded by hand, with digital lab notebooks. “This digital documentation can speed up development projects. They also protect companies in warranty cases and help with potential patent disputes,” explains Dr. Jörg Bretschneider, who supervised the “DiWan” project at Fraunhofer IWS. “That’s why digital lab notebooks are currently a major trend topic.”

In the future, these digital documentations can be further developed into AI-supported assistance and expert systems. The aim is to create digital knowledge repositories that materials researchers can use and expand. Such continuously expandable digital knowledge representations will take materials research and application to a new level.
 

Digital Assistance and Expert Systems Belong in the Curriculum

In practice, however, it makes sense not to proceed with the second step before the first – as DiWan has clarified. The introduction of digital lab notebooks should be prepared, carried out modularly, discussed with employees, and accompanied by further training. New working environments and special care in the structured recording and description of material data also need to be implemented. In addition, future assistance and expert systems must record data sets as completely as possible from the outset, be they text documents, empirical knowledge, measurement data, or metadata. Only a complete collection of all data will pave the way for future AI-supported data exploration. The project partners also advocate using digital assistance and expert systems in the curriculum for skilled worker training. At the same time, they warn against exaggerated expectations: Integrating digital tools into everyday laboratory work is time-consuming. As a result, the personnel costs in the laboratories may initially be higher than before.
 

DiWan Findings Contributed to “Winter School”

The Fraunhofer experts want to incorporate these and other experiences into their consulting services. Digital lab books currently available for materials science and engineering were already presented during the “Winter School” of the “National Research Data Infrastructure for Materials Science & Engineering” (NFDI MatWerk) and the Research Training Group (RTG) 2868 of the Technische Universität Dresden. Competence area manager Martina Zimmermann is also contributing her experience from the DiWan project to the ongoing activities of the NFDI MatWerk. She acts as co-spokesperson in the “Task Area Community Interaction” working group.

In addition to the Fraunhofer IWS, the Dresden University of Technology, imq-Ingenieurbetrieb für Materialprüfung, Qualitätssicherung und Schweißtechnik GmbH, IMA Materialforschung und Anwendungstechnik GmbH (Applus+ IMA), the testing machine manufacturer Hegewald und Peschke Meß- und Prüftechnik GmbH and the software developer Labforward were involved in the DiWan project itself.