May 19 2021
Cornell University is partnering in a $36 million grant from the Toyota Research Institute (TRI) for its Accelerated Materials Design and Discovery (AMDD) collaborative university research program, which seeks to use artificial intelligence to discover new materials that could help achieve emissions-free driving.
Carla Gomes, the Ronald C. and Antonia V. Nielsen Professor of Computing and Information Science, is among the lead researchers on the four-year, multi-institution grant.
"AI for scientific discovery is among the most promising but challenging areas of AI," said Gomes, who is also director of Cornell's Institute for Computational Sustainability. "TRI's multidisciplinary approach enables our team to develop new AI approaches for materials discovery. We combine data-driven deep learning with knowledge-driven reasoning and optimization techniques. This approach allows us to inject scientific background knowledge into the discovery and data-analysis process."
AMDD launched in 2017 with the aim of using AI to discover new materials for emissions-free mobility. The total scope of the initial investment was $35 million over four years, across multiple university partners. The new round of funding adds to the initial investment and seeks ongoing collaboration with program partners, including Cornell.
"Our focus on collaboration is what makes our research program unique," said Brian Storey, director of TRI's AMDD program. "Rather than acting strictly as a funding source, TRI has formed deep collaborations with researchers which have led to joint publications as well as co-developed open-source data and software. This collaborative approach is critical to accelerate the development of new materials for battery and fuel cell vehicles, as no single entity can do this alone."
As part of this initiative, Gomes will collaborate closely with John Gregoire, a staff scientist at the California Institute of Technology. The research focuses on using AI for accelerating high-throughput experimentation for materials discovery, and in particular the discovery of new clean energy materials.
"Our research has led to fundamentally new ways of using AI and machine learning methods to explore the vastness of material space," Gomes said. "For example, you can create all kinds of materials by combining elements, such as the ones we find in the periodic table. Part of the issue is which elements you should combine for obtaining certain properties, like semiconductors or catalysts. This exploration involves discovering the combinations of elements and what the synthesis conditions should be."
With the grant, Gomes and collaborators in Cornell Bowers CIS and the College of Engineering will explore material space from hypothesis formulation to the planning, design and execution of experiments, using semi-autonomous and eventually autonomous systems. If the researchers are successful, Gomes said, at some point AI could make decisions on the discovery of new materials that will contribute to the efficient and beneficial use of Earth's resources.