[Osaka Institute of Technology] 9.5% Adoption Rate for AI for Science Pioneering Challenge Research Creation Project

Osaka Institute of Technology (President: Susumu Inoue) announced that 4 out of 42 research proposals submitted were adopted in the first call for the "AI for Science Pioneering Challenge Research Creation Project (SPReAD)" under the "AI for Science Scientific Research Innovation Program" implemented by the Ministry of Education, Culture, Sports, Science and Technology for Reiwa 8 (2026).

The adoption rate of 9.5% significantly exceeded the overall average of 2.9% and ranked first among universities and institutions that applied for 40 or more projects.

Key Points:

* 4 projects adopted in the first call for the "AI for Science Pioneering Challenge Research Creation Project" * Adoption rate of 9.5% ranked first among universities and institutions applying for 40 or more projects * 2 of the 61 adopted student research projects nationwide are from our university's graduate students

SPReAD is a project by the Ministry of Education, Culture, Sports, Science and Technology that supports challenges with new ideas leading to the advancement and acceleration of scientific research through the utilization of AI. It began this fiscal year against the backdrop of rapid AI advancements and societal penetration in recent years. The program is open to university researchers, students, and private researchers, and covers all fields. The first call for proposals received 15,868 applications from 787 universities and institutions nationwide, with 456 projects adopted.

Our university submitted 42 proposals and had 4 adopted. The 4 adopted projects rank third among private universities nationwide and first among private universities in the Kansai region (2 prefectures and 4 prefectures). Furthermore, 2,624 proposals were submitted by students nationwide, with 61 adopted. Of these, 2 were from our university's graduate students.

Regarding these results, our university believes that its efforts to promote interdisciplinary collaboration and create a research environment that fuses AI and information technology with specialized expertise in each field have borne fruit. We will continue to further strengthen our research system and deepen collaborations to promote innovative achievements and their societal implementation.

The adopted research projects from our university (research representative, title, and abstract) are as follows:

1. Haruna Takagi, 2nd year Master's student, Department of Electrical and Electronic Engineering and Mechanical Engineering, Graduate School of Engineering

"Development of a Non-Contact Measurement Platform for Quantitative Magnetic Field Data from Magneto-Optical Images Using AI"

This project aims to develop an AI-assisted non-contact measurement platform that converts magneto-optical imaging (MOI) images, where it is difficult to directly read the magnitude and direction of magnetic fields, into quantitative magnetic field data based on a self-developed deep learning model, thereby advancing MOI from observation to measurement.

2. Naoya Okumura, 1st year Doctoral student, same department, Robotics and Design Engineering Graduate School

"Acquisition of Gait for Morphing Legged Robots via Sim2Real"

For robots capable of changing between wheeled and legged forms, this project utilizes AI technology to achieve stable movement control on steps and stairs by applying reinforcement learning to robot posture control and walking motion within a physics simulator.

3. Toshiji Nishiguchi, Professor, Department of Real-World Information, Faculty of Information Science and Technology

"Creation of Real-World Information Representation Supporting Safe Movement for Humans and Robots through VLM/RAG and SAT/ILP Verification"

This project will construct a small-scale proof of concept that connects 3D point clouds and color images observed by depth cameras with linguistic knowledge from standards and laws using vision-language models and retrieval-augmented generation, enabling visually impaired individuals and robots to determine safe directions of travel.

4. Toru Ochi, Associate Professor, Department of Information Intelligence, Faculty of Information Science and Technology

"AI Analysis of Good-Faith Misinformation Diffusion Under Social Anxiety"

This project targets SNS information prone to being linked with social anxiety, such as in healthcare, disaster response, and crime prevention. It will use generative AI to analyze the process of "good-faith misinformation diffusion," where general users unknowingly share misinformation believing it to be "useful information."