$1.2 million for two new Stanford research projects on energy/climate AI and environmental justice
Three Stanford University entities will fund two new research projects on using artificial intelligence and machine learning to make energy systems more sustainable, affordable, resilient and fair to all socioeconomic groups.
The projects – funded by Stanford’s Precourt Institute for Energy, the Stanford Institute for Human-Centered Artificial Intelligence (HAI), and the Bits & Watts Initiative – are the first two Precourt Pioneering Projects. The new program aims to fund one new project led by a Stanford faculty member every quarter at a level greater than that provided through the institute’s seed grant program. However, in its first round, leaders of the three entities decided to support two related projects. The resulting tools and datasets from both projects will be made available to researchers beyond Stanford.
“Both research teams proposed really exciting ideas for using massive data to transition our energy system to meet multiple goals simultaneously,” said Yi Cui, director of the Precourt Institute, “so we decided to support both projects.”
“In addition to optimizing for climate change, cost and reliability, they incorporate environmental justice and social equity criteria, which Stanford is committed to,” said Cui, who is also a professor of materials science in the School of Engineering and of photon science at SLAC National Accelerator Laboratory.
SLAC, a U.S. Department of Energy national lab operated by Stanford, will coordinate with Precourt to fund researchers in these two broad research directions. This aligns with the climate and energy research priorities of the current U.S. administration.
“We are excited to partner with Precourt and deepen our existing linkages,” said SLAC Director Chi-Chang Kao, who is also a professor of photon science. “SLAC has thriving efforts in machine learning and applied energy, and the lab is dedicated to advancing environmental justice and equity.”
As Stanford moves to create a new school on climate and sustainability, the leaders of the four entities involved hope that the two new projects and related Stanford research will help recruit new faculty, bridge sustainability research across campus, and attract students of the highest caliber.
“At HAI, we believe that artificial intelligence has the power to help with some of the biggest challenges of our time,” said HAI’s Denning Co-Director Fei-Fei Li, who is also a professor of computer science. “Climate and energy certainly top the list of earth’s most urgent issues. It’s truly our pleasure to support the Precourt Institute for Energy’s Precourt Pioneering Projects grant awards.”
Energy and climate AI hub
One project will build a platform – MESMERIZE: A Macro-Energy System Model with Equity, Realism and Insight in Zero Emissions – centered on how policies and people shape the needed transition to sustainable energy systems and its distributional/equity consequences. The hub will integrate a modelling effort, data sets, advanced computational algorithms and other tools developed at Stanford to solve energy and climate challenges to deep decarbonization.
The project team will use the hub to build a multidisciplinary, economy-wide decarbonization model that integrates social equity and human health concerns. The platform will be a resource for researchers at Stanford and elsewhere to identify and optimize the most effective technological, financial and equitable solutions for different U.S. regions and energy sectors, including electricity, natural gas, transportation and heating.
“The question we want to address is: What are realistic and implementable pathways for sustainable and deeply decarbonized energy systems that include features of real policies, people’s decisions and behaviors, and account for environmental justice?,” said Ines Azevedo, associate professor in the Department of Energy Resources Engineering in Stanford’s School of Earth, Energy & Environmental Sciences.
“We want this interdisciplinary simulation and optimization modeling hub to provide resources to others,” said Azevedo, whose co-leaders on the project are professors Sally Benson, Adam Brandt, Ram Rajagopal and John Weyant, as well as visiting scholar Jacques de Chalendar. “We hope to catalyze more efficient and effective collaborations across campus and beyond by lowering the barriers to sharing knowledge, data, methods and analytical tools.”
Making infrastructure more adaptive
The other project will build open-source tools to assess, forecast and plan for a human-centered infrastructure system with a particular focus on electricity to meet these criteria: decarbonization, equity, affordability and resiliency to the impacts of climate change, including extreme weather events. The research team, led by professors Ram Rajagopal, Arun Majumdar, and Azevedo, as well as adjunct professor Andrew Ng, will use machine learning and publicly available data sources. Other approaches using machine learning do not optimize those four criteria simultaneously.
“The electricity grid is being transformed due to the urgency to decarbonize, improve resilience against climate-induced extreme weather events, and provide affordable, reliable access to at-risk communities,” said Rajagopal, who is an associate professor in the Department of Civil & Environmental Engineering.
“The combination of rapid adoption of renewables, electric vehicles, heat pumps for residential heating and natural gas generation as a transition technology are creating deep interactions among the power grid, natural gas, transportation and information,” Rajagopal explained.
The project will develop three tools that enable granular, interconnected analysis of access, reliability, cost and emissions. The first will assess and predict the risks from climate-related extreme events to local communities, and produce climate-risk scores for communities. These risks include energy insecurity, ill health and other social impacts, particularly as they affect vulnerable populations. The second tool will use remotely sensed data and artificial intelligence to create detailed, high-resolution mapping of U.S. energy resources and infrastructure. Stanford researchers have already used this technology to map specific facets of U.S. energy infrastructure. The third tool will evaluate dynamics in demand and supply due to changing grid conditions: from short-term shocks like extreme weather, to longer term transformations like increased adoption of residential solar power.
The research team will share their data with other researchers as an energy data commons on datacommons.org.
SLAC, operated by Stanford for the U.S. Department of Energy’s Office of Science, explores how the universe works at the biggest, smallest and fastest scales and invents powerful tools used by scientists around the globe. HAI's mission is to advance AI research, education, policy and practice to improve the human condition. The Precourt Institute for Energy is a cross-campus research and education program to make energy more sustainable, affordable and secure for all people. The Bits & Watts Initiative, a Precourt Institute program, finds innovative solutions to power the 21st century electric grid.
The second annual Trans-Pacific Sustainability Dialogue met in Seoul to generate new research and policy collaborations to propel energy security and sustainable energy solutions.
"All of us in science are thinking about how we can take climate and clean energy research to the next level." –John Sarrao
Computer vision is an essential technique that can help policymakers understand residential solar usage.