Chen, Ya-FanYa-FanChenArendt, FelixFelixArendtBornhöft, HansjörgHansjörgBornhöftde Camargo, Andréa S.S.Andréa S.S.de CamargoDeubener, JoachimJoachimDeubenerDiegeler, AndreasAndreasDiegelerGogula, ShravyaShravyaGogulaContreras Jaimes, Altair-TeresaAltair-TeresaContreras JaimesKempf, SebastianSebastianKempfKilo, MartinMartinKiloLimbach, RenéRenéLimbachMüller, RalfRalfMüllerNiebergall, RickRickNiebergallPan, ZhiwenZhiwenPanPuppe, FrankFrankPuppeReinsch, StefanStefanReinschSchottner, GerhardGerhardSchottnerStier, SimonSimonStierWaurischk, TinaTinaWaurischkWondraczek, LotharLotharWondraczekSierka, MarekMarekSierka2025-03-202025-03-202025https://publica.fraunhofer.de/handle/publica/48569610.1002/adem.202401560The development of new glasses is often hampered by inefficient trial-and-error approaches. The traditional glass manufacturing process is not only time-consuming, but also difficult to reproduce with inevitable variations in process parameters. These challenges are addressed by implementing an ontology-based digital infrastructure coupled with a robotic melting system. This system facilitates high-throughput glass synthesis and ensures the collection of consistent process data. In addition, the digital infrastructure includes machine learning models for predicting glass properties and a tool for extracting patent information. Current glass databases have significant gaps in the relationships between compositions, process parameters, and properties due to inconsistent studies and nonconforming units. In addition, process parameters are often omitted, and even original literature references provide limited information. By continuously expanding the database with consistent, high-quality data, it is aimed to fill these gaps and accelerate the glass development process.enapplication ontologydigital infrastructureglass ontologyglassessemantic data integration500 Naturwissenschaften und Mathematik::540 Chemie500 Naturwissenschaften und Mathematik::530 Physikjournal article