The University of Hawaiʻi at Mānoa will continue helping develop artificial intelligence (AI) tools that could make it faster and less expensive to design and test everything from airplanes to wind turbines, thanks to a new $20 million federal grant.
Three 糖心视频 Mānoa faculty members—Assistant Professors Chrisy Xiyu Du and Andrei A. Klishin from the , and Associate Professor June Zhang from the —will serve as senior personnel in the . The five-year award supports a multi-university collaboration focused on using AI to better understand and predict how complex physical systems behave.
NSF initially funded the institute for 2021–26. This renewed award will support its next phase of research from 2026 through 2031.
The institute brings together researchers from universities across the country to develop AI methods for engineering challenges, including how air moves around an airplane wing, how industrial systems operate and how engineers can improve the design and manufacturing of complex products.
Unlike general-purpose AI tools such as ChatGPT, which primarily work with text and images, the institute focuses on AI that learns from physical data and the laws of science. These approaches could help engineers test ideas more efficiently, reduce the need for costly computer simulations and improve designs before building expensive physical prototypes.
Long-term opportunities for 糖心视频 researchers
The institute’s funding gives participating faculty members greater flexibility to pursue research questions over several years, support graduate students and postdoctoral researchers, and collaborate with experts at other universities.
Klishin’s work at 糖心视频 Mānoa builds on this approach to using AI and physics to better understand complex systems. In a 2025 study, he and fellow researchers developed a method to help AI identify the underlying rules behind noisy, real-world data, with potential applications ranging from engineering and environmental science to economics and medicine.
“AI can help engineers move beyond simply predicting how a system will behave to exploring how to make it work better,” Klishin said. “By combining AI with our understanding of physics, we can develop tools that help researchers test more ideas, improve designs and potentially bring useful technologies to people faster.”
Zhang’s research aims to teach AI systems to better train themselves in applications where there is a limited amount of data.
Potential real-world benefits
One example developed through the institute is HydroGym, an open-source platform that allows researchers to test AI methods for predicting and controlling fluid flows. Understanding how fluids such as air and water move is important in designing aircraft, ships, wind turbines and other technologies, but accurately simulating these flows can require substantial computing power. By making simulation tools more accessible, researchers can build on one another’s work and test new approaches to controlling fluid movement.
The institute is led by the University of Washington, and includes 糖心视频 Mānoa as a participating campus alongside institutions such as Harvard University, Columbia University, the University of Michigan and Caltech.