Generative artificial intelligence models can already design unprecedented atomic structures, but they remain unable to determine how to synthesise them: the experimental parameters still have to be guessed.
The ERC Synergy–funded GEMPROMISE project aims to tackle this challenge by bringing together three leading areas of expertise: materials modeling (Gian‑Marco Rignanese, WEL Research Institute, UCLouvain), process chemistry (Cyril Aymonier, CNRS Bordeaux), and generative AI (Pierre Vandergheynst, EPFL Lausanne). Their goal is to develop a model capable not only of imagining new materials, but also of proposing the optimal synthesis conditions for each of them.
This ambitious and deeply collaborative approach could transform the creation of innovative materials.
