UC Berkeley, Georgia Tech and USC launch new National AI Research Institute

The National Science Foundation has awarded $20 million over five years to the Georgia Institute of Technology, the University of California, Berkeley and the University of Southern California to establish the National Artificial Intelligence (AI) Institute for Advances in Optimization. The award is among 11 new National AI Research Institutes announced by NSF.

The AI Institute for Advances in Optimization aims to deliver a paradigm shift in automated decision-making at massive scales by fusing AI and mathematical optimization, to achieve breakthroughs that neither field can achieve independently.

“The frequent power outages in response to wildfires in the Western United States highlighted the importance of resilient distribution of energy,” said Alper Atamturk, the UC Berkeley lead for the new institute and chair of the Department of Industrial Engineering and Operations Research (IEOR). “We also saw the challenges in supply chains early in the pandemic when PPEs and medical equipment could not be delivered to hospitals. By unifying data-driven and model-driven approaches at the core of AI and operations research, the institute will help deliver the next generation of control and optimization algorithms for operating electricity grids with distributed renewable generation, and for designing and operating efficient and resilient supply chains.”

Pascal Van Hentenryck, professor and associate chair for Innovation and Entrepreneurship at Georgia Tech’s School of Industrial and Systems Engineering, is the lead principal investigator of the AI institute. The institute includes 10 UC Berkeley faculty members, including co-principal investigator Dorit Hochbaum, professor in the IEOR department.

“UC Berkeley is a leader in AI, machine learning and optimization,” said Atamturk. “We are delighted to join forces with interdisciplinary research teams at Georgia Tech and USC to advance AI and optimization technologies to address grand challenges in highly constrained settings such as logistics and supply chains, energy and sustainability, and circuit design and control.”

To transform decision-making at massive scales, the institute will develop approaches that not only predict and quantify uncertainty, but also continuously learn and reason to develop the best solutions for the challenges at hand. The AI institute’s methodology thrusts include a new generation of algorithms that use end-to-end learning to tightly integrate forecasting and decision-making, and novel machine-learning methods that can handle massive combinations of factors and constraints.

A key element of the Institute is a partnership with historically Black colleges and universities (HBCUs) in Georgia and Hispanic-serving community colleges in California to build longitudinal education and workforce development programs. The institute will support the development of new AI education and research programs in these colleges and universities, addressing the widening gap in job opportunities in technology fields.

Source: UC Berkeley