Facebook has announced a collaboration with Carnegie Mellon University on a new project called the Open Catalyst Project, whose goal is to leverage advances in AI towards the discovery of new electrocatalysts which will help store and use renewable energy more efficiently.
As Facebook AI’s research scientist Larry Zitnick points out in the announcement blog post, renewable energy sources such as wind and solar provide only intermittent power supply, making it difficult for renewable energy to scale and cover the energy needs of our cities. One scalable solution to this problem of mismatch between the needs and renewable energy production is to store this energy converted into other fuels. These conversion reactions, in turn, need so-called electrocatalysts in order for them to happen.
As we haven’t discovered yet a cheap way to find these electrocatalysts, researchers often rely on heavy computations, which can only provide a limited number of experiments and therefore cannot explore even a small portion of the large search space.
With the novel project Open Catalyst Project, researchers from Facebook AI and Carnegie Mellon University hope to find a way to use AI and find electrocatalysts in a faster and more efficient way. In fact, they hope to utilize AI to approximate quantum mechanical simulations that aim to simulate the movements of atoms in particular reactions.
As part of the project, researchers open-sourced the Open Catalyst 2020 dataset, together with some baseline models and are inviting the research community to take part and help the fight against climate change.