NSF IRES: Physical AI Design

This NSF International Research Experiences for Students (IRES) project in Physical AI Design aims to sponsor U.S. students for summer research internships. The focus will be on developing physics simulation algorithms, user interfaces, and AI algorithms at The University of Tokyo and Japan Advanced Institute of Science and Technology (JAIST). Funded by the NSF Office of International Science and Engineering, the project is led by Prof. Bo Zhu at Georgia Tech. It is conducted in collaboration with Prof. Takeo Igarashi at The University of Tokyo and Prof. Haoran Xie at JAIST. The research team is focused on developing fast simulation and ML algorithms, interactive design user interfaces, computer graphics applications, and an integrated open-source codebase. Our goal is to create novel embodied AI algorithms for complex physical systems. We have a particular interest in devising innovative AI algorithms and user interfaces to facilitate the simulation, design, and optimization of various physical systems. These include soft and rigid bodies, shells and cloth, fluids, and multi-physics coupling systems. Additionally, we aim to develop AI-based interactive systems that enable novel interactions and control between human users and physical objects. Concurrently, the IRES interns will collaborate with students from Georgia Tech, The University of Tokyo, and JAIST to develop an open-source platform for creating accessible, interactive physical AI designs.

Each IRES intern student will be advised jointly by Prof. Bo Zhu at Georgia Tech, Prof. Takeo Igarashi at the University of Tokyo, and Prof. Haoran Xie at Japan Advanced Institute of Science and Technology (JAIST). The three advisors have broad research expertise in computer graphics, physics-based simulation, human-computer interaction, user-design interface, and machine learning.