TFC × TEL × RMC Collaborative Project: Sustainable Open Innovation Seminar
— AI for Science: Exploring the Frontiers of Society and Knowledge —
Special Session 1
Information
Date
Friday, October 23, 2026
Venue
- TOKYO ELECTRON House of Creativity 3F, Lecture Theater, Katahira Campus, Tohoku University [Access]
- Online
Overview of the Seminar Series
In recent years, AI (Artificial Intelligence) has attracted significant attention as a powerful tool for addressing societal challenges. At the same time, it is rapidly becoming a foundational technology that is transforming the very nature of academic research itself, across diverse fields such as materials science, life sciences, linguistics, and space science.
The purpose of this seminar series is to reconsider AI as a “new foundation for scientific inquiry” that deepens our understanding of science. Specifically, the seminar will introduce cutting—edge research from three perspectives:
(1) AI research that aims to autonomously perform scientific reasoning,
(2) AI for science that accelerates simulation and analysis, and
(3) applied AI research that connects these advances to real—world problem solving, such as materials development and drug discovery.
Through interdisciplinary discussions, the series aims to illuminate the paradigm shifts enabled by AI and its potential contributions to a sustainable society. It will also explore how AI expands the ways researchers think, explore, and generate new knowledge.
Speaker
Frank Wang (Research Director, NVIDIA)
Registration
Registration deadline: Monday, October 19, 2026 17:00 (JST)
Registration Form
Time Schedule
- Moderator: Lin Gu (Research Institute of Electrical Communication, Tohoku University)
- 14:00 – 15:00
- Frank Wang (Research Director, NVIDIA)
- Beyond Chatbots: Agentic AI for the Real World
Abstract
AI is moving beyond chatbots that answer questions toward agentic systems that can understand the physical world, plan, and act. This talk presents our recent research featured at leading AI conferences including CVPR and NeurIPS, including on spatial reasoning, vision-language-action, and efficient tool use. It shows how AI can move from visual understanding to real-world action at practical scale, while examining the emerging challenges of fairness in visual AI and safety in embodied agents. Beyond these applications, the talk highlights a broader opportunity for universities: agentic AI may become a new partner for AI for Science, accelerating discovery across engineering, the natural sciences, and medicine.
Poster
- Download [PDF]
Hosted by
- Tokyo Electron Limited
- Tohoku Forum for Creativity, Organization for Research Promotion, Tohoku University
- Research Management Center (RMC), Organization for Research Promotion, Tohoku University
Organizers
Yuji Ogino (Tokyo Electron Limited)
Michio Yoshiwaki (Tohoku Forum for Creativity, Tohoku University)
Marc Hansen (Research Management Center, Tohoku University)
Ryo Sugawara (Research Management Center, Tohoku University)
Yoshihiko Kawasaki (Research Management Center, Tohoku University)
Yoshiaki Maeda (Tohoku Forum for Creativity, Tohoku University)
Contact
tfc_se*grp.tohoku.ac.jp (change * to @)