Project GlasDigital

21/12/2021

• Challenge
• Objectives
• Methods
• Partners

Glasses stand out by their wide and continuously tunable chemical composition and the large variety of shaping techniques. The thereby enabled broad spectrum of specific property profiles makes glasses a key component of modern high technologies. However, the current glass development is mostly based on empirical methods. Due to the small degree of automation, glass development cycles are often cost- and energy-intensive and do not comply with the required short development periods.

Aim of the joint project GlasDigital is to elaborate digital tools as part of an ontology-based digital infrastructure for the data driven high-throughput (HD) glass development in Germany. In this context, essential foundations of a robotic, fully automated and intelligent expert system are to be laid.

The development relies on a robotic glass melting system, for which a prototype was developed at Fraunhofer ISC, and exemplarily starts from a technically improved system at BAM. The enhanced front-end tool and the coupled system control software will be expanded in this regard and linked to the platform MaterialDigital (PMD).

Partners
Bundesanstalt für Materialforschung und -prüfung (BAM)
TU Clausthal
Friedrich-Schiller-Universität Jena
Fraunhofer Institut für Silicatforschung Würzburg

Funding
The project is funded by the Federal Ministry of Education and Research (BMBF) as part of the Initiative for the Digitisation of Materials Research in Germany (MaterialDigital).

GlasDigital - Data-driven workflow for accelerated glass development

GlasDigital wants to elaborate digital tools for the high throughput development of novel glasses. Due to the low degree of automation, current glass development cycles are usually very cost, energy and time expensive. Particular challenges arise from the wide variability of possible chemical compositions, high process temperatures and the lack of process and glass composition-structure-property data. Robotic glass melting systems with self-learning machines and their integration into an ontology based digital infrastructure should overcome these problems.

Methodology and BAM's contribution

The development relies on a robotic glass melting system, for which a prototype was developed at Fraunhofer ISC, and exemplarily starts from a technically improved system at BAM. For this purpose, the system has to be equipped with inline sensors for process monitoring, machine learning (ML)-based, adaptive algorithms for melt monitoring and optimization, novel tools for HD glass analysis and ML-based, adaptive algorithms for glass design, including adaptive software tools for data mining. The enhanced front-end tool and the coupled system control software will be expanded in this regard and linked to the platform MaterialDigital (PMD).

Partners

Bundesanstalt für Materialforschung und -prüfung (BAM)
TU Clausthal
Friedrich-Schiller-Universität Jena
Fraunhofer Institut für Silicatforschung Würzburg

Funding

The project is funded by the Federal Ministry of Education and Research BMBF as part of the Initiative for the Digitization of Materials Research in Germany MaterialDigital