Glass production uses a wide array of sensors and cameras that generate vast amounts of data throughout the entire process chain. This data promises new insights and deeper process knowledge. But bringing together the right information from a heterogeneous data landscape and assessing it quickly to apply it effectively to process control can be a challenge. SCHOTT’s “Sensors & Big Data” project with experts from research and the Business Units aims to do just that to develop concrete solutions. “With new sensors and technologies from big data, we can achieve better process stability and higher yield,” explains Rainer Eichholz, Project Manager and Head of Development for Melting and Drawing Processes at SCHOTT’s Mitterteich production site.
The project team has been working on melting tank concepts for tubing production, optical glass, cooktop panels and thin glass. Their next focus will be to structure and connect data landscapes, making all information accessible from one source. They’ll also use high resolution cameras and image processing technologies to access processes that were less accessible in the past.
The project supports automated analysis routines in production to efficiently analyze the data acquired through this process. “Automated data analysis with the help of user-friendly data mining tools is the key. Our process engineers can then recognize complex interdependencies and intervene if necessary,” says data analysis expert Dr. Markus van Ackeren. On project completion in 2019, the new big data solutions will be transferred to other Group melting tanks.