Washington, District of Columbia 2021-07-29 12:51:11 –
July 29, 2021
National Science Foundation Today Announcement 11 new artificial intelligence research institutes, including those led by the University of Washington. These institutions are part of a $ 220 million investment across seven AI research disciplines. Each institution receives about $ 20 million over five years.
NS UW-led AI Institute for Dynamic Systems It focuses on basic AI and machine learning theories, algorithms, and applications for real-time learning and control of complex dynamic systems that describe chaotic situations where conditions are constantly changing and difficult to predict.
“Engineering science is undergoing a revolution supported by machine learning and AI algorithms,” said the director of the institute. J. Nathan Katz, Professor of Applied Mathematics UW. “This institute brings together a world-class team of engineers, scientists and mathematicians with the goal of integrating basic AI development with applications for critical new technology applications.”
Researchers are aware of the basic physical principles behind dynamic systems, including situations such as turbulence and how the body recovers from injury. However, these scenarios often occur on multiple timescales at once and can be a combination of many types of physics, so researchers should understand exactly what is happening. Is difficult.
“My favorite dynamic system is eddy,” said the institute’s deputy director. Steve Brunton, UW Associate Professor of Mechanical Engineering. “We literally live and breathe in hydraulic fluids, as do almost all machines. But because of the multiscale complexity of fluids with a cascade of smaller and smaller vortices, they are idealized and controlled. It is still incredibly difficult to predict what a fluid will do outside of the set settings. “
The institute’s overall goal is to integrate physics-based models with AI and machine learning approaches to develop efficient, explainable solutions for data-facing challenges across science and engineering.
“Our specific question is whether we can develop better machine learning technology by incorporating and implementing known physics such as conservation law and symmetry,” Brunton said. .. “Similarly, in complex systems where physics is only partially known or unknown, such as neuroscience and epidemiology, can we use machine learning to learn the” physics “of these systems? ?? “
In addition to research, the institute will focus on training future researchers in this area throughout the education pipeline. Examples include partnering with high school programs focused on AI-related projects, and actively recruiting and supporting recent college graduates from underrepresented groups, US veterans, and first-generation college students. For example, creating a baccalauréat program. They go to graduate school.
“The institute provides a large open source educational material, including lectures, data and code packages to advance and empower AI,” says Kutz. “It’s important to provide AI ethics training to everyone involved in the institute, and to make this training available to the entire community, AI and the ethics of data, data privacy, and AI algorithms. We will implement a disciplined approach to thinking about the impact on new social concerns about ethical application. “
For this institute, UW has with several regional institutions such as the University of Hawaii at Manoa, Montana State University, University of Nevada Reno, Boise State University, Alaska University Ancollage, Portland State University, and Harvard University. We are affiliated. And Columbia University.
“We are very excited to gather the critical mass of amazing and innovative researchers from all over the United States and really move the needle to the development of machine learning technology for physical and engineering dynamic systems. “I will.” Brunton said. “We also have close ties to industry partners such as Boeing, which provides a great opportunity to focus on important and relevant issues and ensure that our technology is actually in use. “
Additional UW researchers who are part of this institute are principal investigators Critica Manohar, Associate Professor of Mechanical Engineering. Mariam Fazel, Professor of Electrical and Computer Engineering. Daniela Witten, Professor of Biostatistics; David Beck, Associate Professor of Chemical Engineering.
US Senator Patty Murray of D-Wash said: “As the University of Washington and other research institutes in our state continue to lead artificial intelligence, this investment is important to ensure that Washington remains a leader in innovation, research, and scientific achievement. We will continue to fight for such a significant federal investment to move this work forward. “
UW is also a partner institute for the AIEDGE Institute, another newly announced NSF Institute led by The Ohio State University. The institute’s goal is to design a next-generation wireless edge network that is extremely efficient, reliable, robust and secure.
These 11 new AI institutions 7 AI institutions Funded in 2020, these laboratories will be expanded to transform the workforce and decision-making support in Washington State, in addition to UW-led laboratories that include a total of 40 states and Washington DC. Agricultural AI Institute for, or the AgAID Institute led by Washington State University. The institute’s goal is to use AI to address some of the biggest agricultural challenges associated with labor, water, weather and climate change.
“Washington is already a leader in artificial intelligence,” said Senator Maria Cantwell of D-Wash. “From the University of Washington’s Tech Policy Lab, which studies epic challenges in artificial intelligence, to Washington State University’s research in precision agriculture, we are ready for these two grants and more artificial intelligence applications. UW works in the area of complex systems to improve areas such as manufacturing, and WSU works on improving agriculture. “
The AI Institute for Dynamic Systems is partially funded by the US Department of Homeland Security.
UW to lead new NSF institute for using artificial intelligence to understand dynamic systems Source link UW to lead new NSF institute for using artificial intelligence to understand dynamic systems