Artificial intelligence to speed up clean energy
A CIFAR-sponsored report recommends combining AI, robotics and materials sciences for clean energy technologies
New methods in AI and robotics could create “self-driving labs” that automate the discovery of new clean energy materials. That’s one possibility laid out in a report by an international panel of experts, sponsored in part by CIFAR. The report makes recommendations that could cut the time to discovery of useful new materials from 20 years down to one or two.
“At the moment, we’re very much like Edison looking for filaments for his light bulb, testing them one by one … until we find the one that works. This report lays out a road map for methods that will let us quickly discover and design materials with exactly the properties we need,” said Alán Aspuru-Guzik (Harvard University), a senior fellow in CIFAR’s Bio-inspired Solar Energy program and lead author of the report.
Materials are the foundation of most clean energy technologies such as advanced batteries and improved solar cells. But discovering new materials is currently time-consuming and expensive. To determine whether they will be useful, newly discovered molecules are run through a long process of simulation, synthesis, and testing.
The report makes six recommendations that will lead to what it calls materials acceleration platforms (MAPs). The platforms would accelerate the pace of discovery by integrating automated robotic machinery, rapid synthesis and characterization of materials and artificial intelligence.
The platforms would help researchers transition from a largely trial-and-error method of materials discovery to one of “inverse design,” in which materials with desired properties could be easily searched for and developed.
The recommendations were announced today at a ceremony in Mexico City. The report is the result of a September 2017 workshop that brought together 55 researchers from around the world. The workshop was part of Mission Innovation, a partnership of 22 countries and the European Union that aims to accelerate global clean energy development. It was sponsored by CIFAR, along with the Mexican Ministry of Energy (SENER) and the U.S. Department of Energy.
The report names six key areas which will need to be developed to create these materials acceleration platforms. They are:
1) “Self-driving laboratories” that design, perform and interpret experiments in an automated way;
2) The development of specific forms of AI for materials discovery;
3) Modular materials robotics platforms that can be assemblies of modular building blocks for synthesis and characterization;
4) Further research into computational methods for inverse design;
5) New methodologies for bridging the length and timescales associated with materials simulation; and
6) Sophisticated data infrastructure and interchange platforms.
The report emphasized the need to develop multidisciplinary international teams of scientists and engineers with expertise in chemistry, materials sciences, advanced computing, robotics and AI, among other disciplines.
“I’m pleased that CIFAR was able to contribute to Mission Innovation’s important work,” said CIFAR President & CEO Alan Bernstein. “CIFAR and Mission Innovation share similar goals, and our emphasis on excellence, global participation and tackling tough questions is the best strategy for creating the disruptive technologies needed to address the world’s growing demand for energy.”
Read the full report here.
CIFAR announces distinguished advisory committee for $125M Pan-Canadian Artificial Intelligence Strategy
International Scientific Advisory Committee will recommend Canada CIFAR AI Chair appointments and provide advice on the AI Strategy. CIFAR is...
Opinion piece by Ted Sargent, Alán Aspuru-Guzik, Yoshua Bengio and others argues that machine learning could focus the efforts of...
A CIFAR Distinguished Fellow defined this year and we should look to a CIFAR Azrieli Global Scholar for big things...