Formulations typically contain between seven and 15 different components, giving many variables to optimise when creating a formulation. The automated lab will allow more extensive series of tests to be carried out thanks to computer-aided test series design and the use of automation.
It will test formulations using Covestro binders and crosslinkers, allowing properties such as hardness, adhesion, opacity, gloss and durability to be assessed. The lab will also allow formulations incorporating bio-based or recycled raw materials to be checked rapidly, and tests on raw materials to be carried out. Different climates can be simulated to check product performance under real-world application conditions.
A further advantage is the large amount of structured data the lab will generate, giving additional insight into how the different factors affect formulation performance. These data, combined with measurement data from existing studies, can be evaluated using machine learning algorithms to improve formulations further. AI tools can be used to predict new experiments based on property goals and verify them in the automated lab, creating a self-learning system.
“With our automated laboratory, we can work together with our customers on the future of coatings and adhesives,” said Thomas Büsgen, the lab’s head. “Because it operates almost completely autonomously and learns from our existing knowledge and data lake as well as newly generated data, it makes the process of optimising and developing formulations many times more efficient and precise.”