Recently, Linton Crystal Technologies partnered with the Golisano Institute of Sustainability at Rochester Institute of Technology to conduct research on the crystal ingot growth process. Researchers at RIT used process time-series data as well as images from the KICCS cameras integral to Linton’s Czochralski crystal growers to develop data-driven models to predict errors during crystal formation.

The researchers found the results of the image-based model to be very promising at successfully detecting quality precursors in real-time operation.

The full case study of the project is available Machine Learning Research Case Study.