Project Q1

Copolymers offer a vast design space across length scales, chemical compositions and only partially understood structure-processing-property relationships. While this complexity seems like an ideal application for data-driven techniques, their usefulness is currently limited by the tiny amount of reliable, interlinked data.

The need for better curated data is most pronounced for the synthesis and characterization of the copolymers. Even though a lot of relevant data will be generated in COIN, most existing data is hidden in journal articles or hard-to-access places such as historic lab notebooks The ever-growing number of publications, however, makes it difficult to systematically leverage this knowledge for the training of models or the design of experiments. The project Q1 aims to lift the hidden insights from the existing knowledge in the literature and internal sources using data mining techniques and make it accessible and reusable to COIN.