AI 4 Optimization (Credit: Dall-E)
The National Science Foundation (NSF) has renewed the Artificial Intelligence Institute for Advances in Optimization (AI4OPT) through 2031, awarding $20 million to support the institute’s next phase of research at the University of Southern California (USC), the University of California, Berkeley, and lead institution Georgia Tech.
Over the next five years, the renewed institute will continue its mission to fuse AI and Optimization to achieve breakthroughs that could not be delivered independently by the two technologies, and to realize their effectiveness in real-world applications at societal scale.
Building on its first five years of work, the institute’s second phase will focus on four integrated areas, or “thrusts”: two advancing the science of AI and optimization, and two applying those advances to energy systems, supply chains and manufacturing.
AI4OPT aims to bring foundational research closer to real-world applications by developing intelligent systems that can make complex decisions, account for uncertainty and coordinate at massive scales.
Georgia Tech professor Pascal Van Hentenryck will continue to serve as the lead principal investigator, with USC’s Bistra Dilkina and Berkeley’s Alper Atamtürk serving as co-principal investigators.
USC’s role has expanded in AI4OPT’s second phase, with the university receiving $5.6 million – more than double its funding from the first phase – and increasing its team to five researchers.
The USC team brings together an interdisciplinary group of researchers across the USC Viterbi School of Engineering, USC Stevens School of Computing and AI and USC Marshall School of Business.
USC will also bring the Seth Bonder summer camp in computational and data science for engineering to Los Angeles, extending the institute’s educational reach to high school students.
“This renewal allows us to dig deeper into AI and optimization, moving basic science toward real-world applications that make a tangible difference in society,” said Dilkina, the Dr. Allen and Charlotte Ginsburg Early Career Chair in Computer Science, co-director of USC Center for AI in Society, and associate professor of Computer Science and Industrial and Systems Engineering. “With an expanded team and increased support, USC is excited to collaborate with our partners to push boundaries and amplify our collective strengths.”
Dilkina will lead the USC team and the institute’s AI for Optimization research thrust.
For more about the renewal, check out Georgia Tech’s press release and UC Berkeley’s release.
Turning AI and Optimization Into Real-World Solutions
As power grids, supply chains and other critical systems become increasingly complex and unpredictable, decision-makers need to respond to changes faster than traditional optimization methods can allow.
AI4OPT’s renewed research aims to meet that demand by teaching machines to solve complex optimization problems more efficiently.
Led by Dilkina, the AI for Optimization thrust will develop “learning to optimize” approaches and foundational models for discrete optimization. These technologies could enable decision-making systems to respond to rapidly changing conditions in milliseconds — from adapting power grids to managing complex logistics.
The Optimization for AI thrust, led by Atamtürk, will develop new optimization techniques to make AI training more scalable, trustworthy and energy efficient, with the potential to support more robust and interpretable models that can operate on resource-constrained devices.
Led by Georgia Tech’s Chelsea White and Daniel Molzahn, the final two thrusts are application-focused and will put these advances to work.
The Energy Systems thrust will address the increasing complexity of electrical grids as renewable energy and data center demand grow, while Supply Chains and Manufacturing will develop more agile systems for logistics, production and semiconductor manufacturing that can better withstand disruptions caused by disasters, pandemics and other global shocks.
USC’s Expanded AI4OPT Team
USC’s interdisciplinary AI4OPT team reflects the university’s broader commitment to translating AI research into practical solutions and advancing AI through the university’s Presidential Initiatives, helping position USC as a leader in AI.
The renewed USC team features computer science professor Willie Neiswanger, who will focus on the AI for Optimization thrust with an emphasis on uncertainty modeling; industrial and systems engineering professor Andres Gomez, who serves as the USC lead for the Optimization for AI thrust; industrial and systems engineering professor Karmel Shehadeh, who will lead research on agile logistics and drone deployment within the Supply Chains and Manufacturing application area; and data sciences and operations professor Vishal Gupta, who will focus on methodologies at the intersection of AI and optimization and their applications to supply chains.
Neiswanger is an assistant professor of computer science at USC Viterbi and USC Steven’s Thomas Lord Department of Computer Science.
Gomez is an associate professor at USC Viterbi and an Epstein Family Early Career Chair in Industrial and Systems Engineering.
Shehadeh is a Gabilan Assistant Professor of Industrial and Systems Engineering at USC Viterbi’s Daniel J. Epstein Department of Industrial and Systems Engineering
Gupta is a Dean’s Associate Professor of Data Sciences and Operations and Industrial and Systems Engineering, with joint appointments in the Daniel J. Epstein Department of Industrial and Systems Engineering and USC Marshall’s Department of Data Sciences and Operations.
Looking Back at AI4OPT’s First Phase
AI4OPT’s renewal builds on its first phase, which produced more than 250 publications, resulted in 12 technology transfers and created new datasets, including Distributional MIPLIB, advancing the field of optimization.
At USC, the institute helped catalyze the university’s leadership in AI and optimization education, including the launch of the first PhD specialization and certificate in AI plus Optimization (USC’s ORAI program) supported by a National Science Foundation Research Traineeship grant.
For Dilkina, the next phase represents an opportunity to build on those advances and expand their scientific and societal impact.
Published on September 4th, 2026
Last updated on October 21st, 2026