Priya Donti, Climate Change AI
Priya Donti, Executive Director of Climate Change AI, explains multiple ways that machine learning and AI can be used to mitigate climate change. Her work at the intersection of climate change, computer science, and data science led her to co-found Climate Change AI, our partner for the WiDS Datathon 2023 challenge to improve long-term weather forecasting.
An expert in climate change and the optimization of power grids, Priya Donti researches how to use machine learning for forecasting, optimization, and control of power grids to facilitate the integration of renewable energy.
She first became interested in climate change during high school and studied computer science with a focus on environmental analysis as an undergraduate at Harvey Mudd College. After graduation, she spent a year on a Watson Fellowship, learning about different approaches for next-generation power grids in Germany, India, South Korea, Chile, and Japan. She went on to earn her PhD in power grid optimization at Carnegie Mellon. While there, she co-founded Climate Change AI, an initiative born out of a paper she co-wrote with academic and industry leaders about the ways machine learning could address climate change.
Machine learning can play a role in mitigating climate change in areas like decarbonizing power grids, buildings, and transportation; helping create more precise forecasts for climate change impacts; and strengthening social, food, and health systems to cope with the impacts of climate change.
There are several ways to apply machine learning to the climate crisis. One is distilling raw data into actionable insights, like turning satellite imagery into inputs on where the solar panels are or where deforestation might be happening, or turning large amounts of text documents into insights to guide policy or innovation. A second way is forecasting solar and wind power, and extreme weather events. A third is optimizing complex systems to make them more efficient, like heating and cooling systems in buildings or optimizing freight transportation systems. Machine learning is also valuable in science and engineering workflows to accelerate the design of new batteries or speed up climate or power models.
While there are many ways that AI and data science can play a role in climate action, sometimes it’s difficult figuring out where to start. Priya says the WiDS Datathon is a great way to get started because no matter how much experience you have, you can enter and be able to work on this particular challenge. “The floor is low, but the ceiling in high.” There are also many resources on the Climate Change AI website to start learning, get involved, and meet other people working in the space through workshops, virtual happy hours, mentorship programs, and an online community platform.
Connect with Priya on LinkedIN
Find out more about the Climate Change AI
Connect with Margot Gerritsen on Twitter (@margootjeg) and LinkedIn
Find out more about Margot on her Stanford Profile
Listen and Subscribe to the WiDS Podcast on Apple Podcasts, Google Podcasts, Spotify, Stitcher
Subscribe to the WiDS Podcast: