Researchers at Duke University, based in North Carolina, USA, are developing energy-efficient Artificial Intelligence technologies while applying AI to climate forecasting, conservation, water management and disease prediction.
Duke University researchers are exploring ways to reduce the environmental impact of Artificial Intelligence while harnessing the technology to address climate change and other environmental challenges.
The work spans engineering, climate science, conservation and public policy, with researchers developing more energy-efficient AI hardware alongside applications designed to improve climate modelling, weather forecasting and environmental monitoring. Duke is also planning a small GPU centre for 2027 designed to minimise electricity and water consumption and carbon emissions.
Toddi Steelman, Vice President and Vice Provost for Climate and Sustainability at Duke University, said: “Climate change is one of the defining challenges of our time, and AI can be an important part of the solution, but only if we develop it responsibly. Duke’s commitment is to lead in both directions: advancing AI that helps society understand and respond to a changing planet while pioneering the energy-efficient, ethical, and equitable technologies needed to make AI itself more sustainable.”
The university’s research comes as growing AI adoption places additional pressure on electricity infrastructure.
Brian Murray, Director of Duke’s Nicholas Institute for Energy, Environment & Sustainability, said large data centre power demands are expected to increase by around 130% by the end of the decade, with AI responsible for approximately half of the load growth. Duke researchers have investigated how greater flexibility in when large electricity users consume power could help grids accommodate increased demand without immediately requiring extensive new capacity.
Researchers at Duke’s Pratt School of Engineering are meanwhile examining how AI hardware itself could become more efficient. Tania Roy, Associate Professor of Electrical and Computer Engineering, is developing brain-inspired semiconductor devices capable of carrying out AI tasks locally rather than relying on data centres.
The approach could eventually allow devices including voice assistants, autonomous vehicles and smart cameras to perform sophisticated AI processing themselves, reducing the amount of information transmitted to energy-intensive data centres. Other Duke researchers are investigating magnetic random access memory and more efficient robotic AI processing hardware.
AI is also being applied to environmental research. Professor Shineng Hu uses Machine Learning and Deep Learning to improve forecasts of climate extremes, including El Niño events, and investigate climate-related societal impacts.
Duke’s Marine Robotics and Remote Sensing Lab is combining AI with drones and satellites to monitor wildlife and coastal environments. Its systems can detect and count animals, identify species, measure bodies, map habitats and track environmental change.
David Johnston, Director of the Lab, said: “In the past, scientists didn’t have enough data. Now the script is flipping, and we are often data rich but without enough time to go through all the data. AI can help us analyse our data in 20% of the time, and we can rapidly see the results.”
The technology is being used to monitor harbour seals in Alaska’s Glacier Bay National Park and Preserve as their glacial habitat changes.
Elsewhere, Marta Zaniolo, Assistant Professor of Civil and Environmental Engineering, is combining AI with hydrology and climatology to improve water-resource decisions, while researchers led by William Pan have developed an AI system capable of predicting malaria outbreaks months in advance.
The malaria forecasting system analyses temperature, precipitation and satellite imagery and has been validated in Peru, Ecuador, Brazil, Panama and Honduras, while being tested in Colombia.
Duke is also considering the wider governance implications of AI infrastructure through its Deep Tech at Duke Initiative, bringing together researchers, policymakers and industry to examine issues including data centres, power generation, cooling, water consumption and ethical governance.
