In cooperation with the Technical University of Berlin, chemical company BASF is working on an AI-based solution to answer fundamental questions about process and quantum chemistry in the Berlin-based Joint Lab for Machine Learning (BASELEARN) project. The company is funding the project with over €2.5 million, with the aim of leveraging machine learning technology to predict chemical processes through the use of tailor-made mathematical models and algorithms.
BASF has provided the researchers with real process data to assist them in their work. This will help them develop machine learning to predict the solubility of complex mixtures or dyes, as well as the aging processes of catalysts, consequently bringing about concrete industrial benefits. The machine learning model is designed to reliably predict the relevant properties even when a large number of components are used. With its participation in the project, the chemical company, which has already opened joint research laboratories with the Karlsruhe Institute of Technology (KIT) and the University of Heidelberg in the past, is strengthening its close links with science.