TY - CONF A1 - Müller, Ralf T1 - GlassDigital: Digital Infrastructure for Data-Driven High-Throughput Glass Development N2 - Gläser zeichnen sich durch eine breite und kontinuierlich abstimmbare chemische Zusammensetzung sowie einzigartige Formgebungstechniken aus, was sie oft zur Schlüsselkomponente moderner Hochtechnologien macht. Die Glasentwicklung ist jedoch oft noch zu kosten-, zeit- und energieintensiv. Der Einsatz von robotergestützten Schmelzsystemen, eingebettet in eine Ontologie-basierte digitale Umgebung, soll diese Probleme in Zukunft überwinden. Im Rahmen der BMBF Forschungsinitiative MaterialDigital unternimmt das Verbundprojekt GlasDigital „Datengetriebener Workflow für die beschleunigte Entwicklung von Glas“ erste Schritte in diese Richtung. Das Projektkonsortium, an dem das Fraunhofer ISC in Würzburg, die Friedrich-Schiller-Universität Jena (OSIM), die Technische Universität Clausthal (INW) und die Bundesanstalt für Materialforschung und -prüfung (BAM, Fachgruppe Glas) beteiligt sind, will alle wesentlichen Basiskomponenten für eine beschleunigte datengetriebene Glasentwicklung zusammenführen. Zu diesem Zweck wird ein robotergestütztes Hochdurchsatz-Glasschmelzsystem mit neuartigen Inline-Sensoren zur Prozessüberwachung, auf maschinellem Lernen (ML) basierenden adaptiven Algorithmen zur Prozessüberwachung und -optimierung, neuartigen Werkzeugen für die Hochdurchsatz-Glasanalyse sowie ML-basierten Algorithmen zum Glasdesign, Data Mining sowie Eigenschafts- und Prozessmodellierung ausgestattet. Der Vortrag gibt einen Überblick darüber, wie all diese Komponenten miteinander verzahnt sind, und veranschaulicht ihre Nutzbarkeit anhand einiger Beispiele. T2 - HVG-Fortbildungskurs CY - Offenbach, Germany DA - 27.11.2023 KW - Glas KW - Ontology KW - Workflow KW - Simulation KW - Robotic melting PY - 2023 AN - OPUS4-60386 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Müller, Ralf T1 - GlasDigital: Data-driven workflow for accelerated glass development N2 - lasses stand out by their wide and continuously tunable chemical composition and large variety of unique shaping techniques making them a key component of modern high technologies. Glass development, however, is still often too cost-, time- and energy-intensive. The use of robotic melting systems embedded in an ontology-based digital environment is intended to overcome these problems in future. As part of the German research initiative MaterialDigital, the joint project GlasDigital takes first steps in this direction. The project consortium involves the Fraunhofer ISC in Würzburg, the Friedrich Schiller University Jena (OSIM), the Clausthal University of Technology (INW), and the Federal Institute for Materials Research and Testing (BAM, Division Glasses) and aims to combine all main basic components required for accelerated data driven glass development. For this purpose, a robotic high throughput glass melting system is equipped with novel inline sensors for process monitoring, machine learning (ML)-based, adaptive algorithms for process monitoring and optimization, novel tools for high throughput glass analysis and ML-based algorithms for glass design, including software tools for data mining as well as property and process modelling. The talk gives an overview how all these tools are interconnected and illustrates their usability with some examples. T2 - HVG-DGG Fachausschuss I CY - Jena, Germany DA - 03.11.2023 KW - Glass KW - Robotic melting KW - Ontologie KW - Simulation KW - Workflow KW - Data Space PY - 2023 AN - OPUS4-60383 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Müller, Ralf T1 - GlasDigital: Data-driven workflow for accelerated glass development N2 - Glasses stand out by their wide and continuously tunable chemical composition and large variety of unique shaping techniques making them a key component of modern high technologies. Glass development, however, is still often too cost-, time- and energy-intensive. The use of robotic melting systems embedded in an ontology-based digital environment is intended to overcome these problems in future. As part of the German research initiative MaterialDigital, the joint project GlasDigital takes first steps in this direction. The project consortium involves the Fraunhofer ISC in Würzburg, the Friedrich Schiller University Jena (OSIM), the Clausthal University of Technology (INW), and the Federal Institute for Materials Research and Testing (BAM, Division Glasses) and aims to combine all main basic components required for accelerated data driven glass development. For this purpose, a robotic high throughput glass melting system is equipped with novel inline sensors for process monitoring, machine learning (ML)-based, adaptive algorithms for process monitoring and optimization, novel tools for high throughput glass analysis and ML-based algorithms for glass design, including software tools for data mining as well as property and process modelling. The talk gives an overview how all these tools are interconnected and illustrates their usability with some examples. T2 - USTV-DGG joint meeting CY - Orleans, France DA - 22.05.2023 KW - Glass KW - Ontology KW - Data Space KW - Workflow KW - Robotic melting PY - 2023 AN - OPUS4-60372 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER - TY - CONF A1 - Ertural, Christina T1 - Vibe Check via Machine Learning: Testing the Prototype N2 - Vibrational properties play a key role in determining the stability and thermal conductivity behaviour of materials. The quasi- harmonic approximation gives insight into the phononic properties of a compound, but in the established way, i.e. density functional theory based methods, it takes many calculation steps and consumes a lot of resources to arrive at the desired results. Machine learning (ML) trained interatomic potentials (e.g. Gaussian approximation potential, GAP) pose an alternative to the traditional computation way of phonons. We develop a Python code based workflow which combines automation tools like atomate2 with ML to ease providing interactomic potentials for (quantum chemical) computations and databases. T2 - #RSCPoster Twitter conference 2023 CY - Online meeting DA - 28.02.2023 KW - Interatomic potentials KW - Machine learning KW - Phonons KW - Workflow PY - 2023 UR - https://twitter.com/cer5814012/status/1630547004462858240 AN - OPUS4-57059 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER -