AI-powered, self-driving laboratories are accelerating materials research, for example in the search for new catalysts. A team from the Fritz Haber Institute and BASF has now demonstrated that this AI can also provide scientific explanations for its results. This new approach was successfully validated in the industrially important conversion of propane to propylene.
The study shows that AI and automation in the chemical industry do not have to come at the expense of understanding. When used thoughtfully, these technologies can fundamentally transform materials development—moving away from trial and error toward genuine understanding. Ultimately, this will lead to AI serving as an active partner in scientific discovery, rather than merely as an efficient but difficult-to-evaluate tool.
Source and link to the full article:
Fritz Haber Institute of the Max Planck Society (June 2026)