For Priya Donti, childhood trips to India were greater than a possibility to go to clan. The biennial journeys activated in her a motivation that continues to shape her research and her teaching.
Contrasting her family home in Massachusetts, Donti — now the Silverman Family Profession Development Professor within the MIT Department of Electrical Engineering and Computer Science (EECS) and a principal investigator on the MIT Laboratory for Information and Decision Systems — was struck by the disparities in how people live.
“It was very clear to me the extent to which inequity is a rampant issue around the globe,” Donti says. “From a young age, I knew that I definitely wanted to deal with that issue.”
That motivation was further stoked by a highschool biology teacher, who focused his class on climate and sustainability.
“We learned that climate change, this huge, necessary issue, would exacerbate inequity,” Donti says. “That actually stuck with me and put a fireplace in my belly.”
So, when Donti enrolled at Harvey Mudd College, she thought she would direct her energy toward the study of chemistry or materials science to create next-generation solar panels.
Those plans, nevertheless, were jilted. Donti “fell in love” with computer science, after which discovered work by researchers in the UK who were arguing that artificial intelligence and machine learning could be essential to assist integrate renewables into power grids.
“It was the primary time I’d seen those two interests brought together,” she says. “I got hooked and have been working on that topic ever since.”
Pursuing a PhD at Carnegie Mellon University, Donti was capable of design her degree to incorporate computer science and public policy. In her research, she explored the necessity for fundamental algorithms and tools that would manage, at scale, power grids relying heavily on renewables.
“I desired to have a hand in developing those algorithms and power kits by creating latest machine learning techniques grounded in computer science,” she says. “But I desired to make certain that the way in which I used to be doing the work was grounded each within the actual energy systems domain and dealing with people in that domain” to supply what was actually needed.
While Donti was working on her PhD, she co-founded a nonprofit called Climate Change AI. Her objective, she says, was to assist the community of individuals involved in climate and sustainability — “be they computer scientists, academics, practitioners, or policymakers” — to come back together and access resources, connection, and education “to assist them along that journey.”
“Within the climate space,” she says, “you wish experts particularly climate change-related sectors, experts in numerous technical and social science tool kits, problem owners, affected users, policymakers who know the regulations — all of those — to have on-the-ground scalable impact.”
When Donti got here to MIT in September 2023, it was not surprising that she was drawn by its initiatives directing the applying of computer science toward society’s biggest problems, especially the present threat to the health of the planet.
“We’re really excited about where technology has a much longer-horizon impact and the way technology, society, and policy all need to work together,” Donti says. “Technology is just not just one-and-done and monetizable within the context of a yr.”
Her work uses deep learning models to include the physics and hard constraints of electrical power systems that employ renewables for higher forecasting, optimization, and control.
“Machine learning is already really widely used for things like solar energy forecasting, which is a prerequisite to managing and balancing power grids,” she says. “My focus is, how do you improve the algorithms for actually balancing power grids within the face of a spread of time-varying renewables?”
Amongst Donti’s breakthroughs is a promising solution for power grid operators to have the ability to optimize for cost, making an allowance for the actual physical realities of the grid, moderately than counting on approximations. While the answer is just not yet deployed, it appears to work 10 times faster, and way more cheaply, than previous technologies, and has attracted the eye of grid operators.
One other technology she is developing works to supply data that might be utilized in training machine learning systems for power system optimization. On the whole, much data related to the systems is private, either since it is proprietary or due to security concerns. Donti and her research group are working to create synthetic data and benchmarks that, Donti says, “may also help to reveal a few of the underlying problems” in making power systems more efficient.
“The query is,” Donti says, “can we bring our datasets to some extent such that they’re just hard enough to drive progress?”
For her efforts, Donti has been awarded the U.S. Department of Energy Computational Science Graduate Fellowship and the NSF Graduate Research Fellowship. She was recognized as a part of ’s 2021 list of “35 Innovators Under 35” and Vox’s 2023 “Future Perfect 50.”
Next spring, Donti will co-teach a category called AI for Climate Motion with Sara Beery, EECS assistant professor, whose focus is AI for biodiversity and ecosystems, and Abigail Bodner, assistant professor within the departments of EECS and Earth, Atmospheric and Planetary Sciences, whose focus is AI for climate and Earth science.
“We’re all super-excited about it,” Donti says.
Coming to MIT, Donti says, “I knew that there could be an ecosystem of people that really cared, not nearly success metrics like publications and citation counts, but concerning the impact of our work on society.”
