Endowed Research Division:
Nanomaterial Process Data Science

Associate Professor Hashimoto delivered an invited lecture at the Quantum Physics and Nanoscience Seminar at Institute of Science Tokyo.

Lecture Title: AI and Robotics: Opening New Doors in Materials Science

The lecture introduced the development of materials maps integrating experimental and computational data through machine learning, automated MOF synthesis using robotic arms, electric pipettes, and 3D printers, and autonomous data analysis methods using multi-agent systems powered by generative AI.

Associate Professor Hashimoto's team delivered a keynote lecture at IMPRES2025 (4th International Symposium on Powder Metallurgy and Materials Development).

Lecture Title: Mapping Thermoelectric Materials Using Machine Learning on Integrated Computational and Experimental Datasets

The presentation introduced a novel approach for mapping thermoelectric materials by integrating computational and experimental datasets using machine learning techniques.

IMPRES2025 Official Website

As part of our AI utilization in research, we have launched the AI-powered textbook writing project (AI Terakoya).

This project creates comprehensive educational content across five domains of materials science (Materials Informatics, Process Informatics, Machine Learning, Materials Science, Fundamentals) using AI.

AI Terakoya: Materials Science Textbook Project

Associate Professor Hashimoto presented research at the Japan Society of Applied Physics (JSAP) Autumn Meeting.

Presentation Title: Application of a Material Structural Similarity Map to Materials Process Exploration

The presentation introduced a novel method for visualizing material structural similarity to enable efficient exploration of materials processes.

Associate Professor Hashimoto and research team published a paper on creating "materials maps" using AI in the international journal APL Machine Learning.

This research developed a materials map integrating experimental and theoretical calculation data using message-passing graph neural networks. The map visually represents structural similarity of materials, enabling rapid extraction of unknown high-performance materials.

View Paper Details

Associate Professor Hashimoto delivered an invited lecture at the GLAM Special Seminar at Stanford University.

Lecture Title: Local and Global Mapping of Thermoelectric Materials Based on Computational and Experimental Datasets

The lecture presented local and global mapping methods for thermoelectric materials based on computational and experimental datasets. Using machine learning-based materials mapping technology, the approach visualizes structure-property correlations of thermoelectric materials and introduces a new method for efficient exploration of high-performance materials.

Associate Professor Hashimoto gave a presentation on experimental automation using robotic arms, invited by TechShare Inc.

The presentation introduced examples of experimental automation systems developed in our laboratory and low-cost automation methods.

We are developing an automated experimental system combining robotic arms (approximately 200,000 yen) and electric pipettes.

The initial system features a high-throughput mixing system that has improved experimental efficiency and reproducibility. This low-cost automation solution is expected to be widely applicable.

We held an inaugural symposium for the Endowed Research Division at FRIS.

The symposium presented the establishment purpose and future research direction of this division, discussing the creation of materials process data science as a new interdisciplinary field. The importance of industry-academia collaboration was confirmed, and active discussions were held on future joint research directions.

Successfully demonstrated automated drawing of the Tohoku University logo using robotic arms.

This experiment demonstrated improved precision control technology of robotic arms and the potential for automating complex experimental operations. This is an important achievement toward advancing experimental automation systems.