Nurturing talents and professionals for the digital age
Results of the FY2024 MITOU Advanced Program
Release Date:Dec 19, 2025
In-page Links
Results of the FY2024 MITOU Advanced Program
PM: URUSHIBARA Shigeru
Digital Twinning and the Science of Geometric Beauty of BONSAI
The "BONSAIENCE" project aims to create a high-accuracy digital archive and pass down bonsai culture by utilizing cutting-edge 3D reconstruction technology to address issues such as craftsmanship skills decline and the threat of theft. The project is built upon three pillars: (1) "BONSAIENCE Museum," a 3D digital archive for the public museum's collection of Omiya Bonsai Art Museum in Saitama, the sacred ground of bonsai history, (2) "BONSAIENCE Hub," a dedicated bonsai digital platform, and (3) "BONSAIENCE Studies," scientific research into bonsai aesthetics. Particularly, BONSAIENCE Studies has undertaken the challenge of developing a bonsai fingerprinting system through the digitization of tree structure data, creating individual bonsai identification technology that can serve as proof of ownership in cases of theft and illegal resale.
Creators: UTSUMI Shinobu, KOYAMA Kenshin, JONO Hitoshi, SHIN Soichiro, ISHIMARU Taisei, OKADA Ken, MATSUOKA Shinryu
Development of a Generative AI-based 3D Printing System
This project developed a system utilizing generative AI and 3D printing technology to efficiently produce customized products tailored to user preferences. Focusing on jewelry, the system employs an image-selection-based AI learning mechanism, allowing customers to simply select images to provide preference data, which automatically generates 3D models. Compared to traditional customization methods, this technology significantly reduces time and costs while offering personalized products quickly. Additionally, it achieved the production of jewelry with unique visual effects, securing a competitive edge. The technology is not limited to jewelry and has potential applications in diverse fields such as furniture and architecture, promising innovation across the manufacturing industry. Moving forward, the project plans to expand its impact by offering SaaS-based services to other companies, driving broader adoption and business growth.
Creators: OBINATA Shin, HORIKAWA Junichiro
Development of New Control Systems for Industrial Freezers as Grid-Interactive Thermal Energy Storage
As electricity costs rise and renewable curtailment persists, the project focuses on the intersection of the power grid and the cost-pressured cold chain, developing control software that enables cold-storage warehouses to act as grid-interactive flexible resources.
A proprietary algorithm ingests each site’s operational and thermal data alongside grid signals, computing freezer duty cycles that pre-cool ahead of expected peaks and shift load away from high-price hours, while strictly honoring temperature and product-safety limits.
In 2024, the project progressed from household-scale prototypes to system architectures suited for large industrial facilities. Simulations using real operational data projected up to ~26% electricity-bill reduction.
The project also established relationships with cold-storage operators and utilities toward future pilot partnerships to enhance grid stability by leveraging existing assets in the cold chain sector.
Creators: MIZUNO Shunsuke, IZUMI Hinata, AKIYAMA Kimihiro, HATA Tamami
PM: KAJITA Mami
Developing an AI Service to Automatically Convert Text into Manga
"Mazin Studio" is a next-generation painting tool that uses generative AI and image generation technology to automatically create high-quality manga from text and sketches as a cloud service. This tool can reduce the labor required for drawing processes by up to 70% compared to traditional hand-drawn methods, offering significant efficiency gains, especially in Webtoon production. Users simply input rough sketches, and the AI interprets intentions to refine the artwork, leveraging a "brush-up loop" process to enhance quality. The "irreversibility" of AI-based workflows and the "innovation dilemma" faced by major painting tool companies ensure competitive advantages for Mazin Studio. This technology is accessible to both professional manga artists and beginners, boosting productivity in manga creation and increasing the number of new entrants, potentially raising market supply significantly.
Creators: HIRUMA Daichi
PM: SHUDO Kazuyuki
Development of general-purpose slicer for multi-axis 3D printer
The "MAGE (Multi Axis Generative Engine)" project has developed affordable multi-axis 3D printers and a universal slicer, addressing the high entry barriers in the market dominated by expensive robotic arms. The 5-axis 3D printer (MAGE Printer) utilizes posture control to eliminate the need for support materials and infills while enabling the creation of complex curved surfaces. The corresponding slicer (MAGE Slicer), employing NURBS surface slicing techniques, has been released as open-source software, reducing barriers for researchers. Additionally, a simulator was developed to preemptively verify potential collisions among components during printing. This technology provides an accessible environment for researchers to engage in multi-axis 3D printer applications, fostering new market creation and application development.
Creators: TAMBO Noriaki
Development of a high-precision physical model-based flight controller platform
"Tobas" is a new flight controller platform designed for industrial and research applications. It addresses the limitations of conventional flight controllers, which rely on specific airframe shapes, by offering physics model-based control that reflects user designs. Key features include the "URDF Builder" for airframe design, "Setup Assistant" for parameter configuration, "Parameter Tuning" for online adjustments, "Flight Log" for flight data management, and "Simulation" for managing simulation environments. Tobas utilizes ROS2 and Raspberry Pi 5 for high scalability and employs Bidirectional DSHOT for precise attitude control. Open-source software (OSS) integration enables community development and aims to expand its user base. This technology enhances design flexibility and efficient drone operations, potentially revitalizing the entire drone industry. A closed beta test is scheduled to begin in 2025.
Creators: DOHI Masayoshi, UETAKE Sora
PM: HARADA Tatsuya
Development of a Digital Twin for Shiitake Mushroom Fields as a Cultivation Management Support System
This project has developed a cultivation management support system for bed-grown shiitake mushrooms. The system uses an autonomous robot to capture video of the cultivation area, processes it in the cloud to generate 3D models, and quantifies mushroom growth. By analyzing per-bed changes in mushroom volume over time, the system optimizes cultivation timing decisions using AI. Unlike traditional smart farming, which mainly focuses on environmental sensing, this system leverages trait-based data analysis, providing a technological edge. The technology is planned to be offered as a SaaS model, encouraging new farmer entry and enabling application to other crops, thereby revitalizing agriculture as a whole. The project aims to enhance the quality and market competitiveness of Japanese-grown shiitake mushrooms while driving innovation in cultivation techniques.
Creators: ISHIZUKA Tatsuya, SASAKI Yusuke, MATSUI Masahiro
Development of Energy-Saving Building Design Support Services Using Combinatorial Optimization and Building Information Modeling
"ZEBopt" is a web application designed to streamline energy-efficient building design by utilizing combinatorial optimization and BIM (Building Information Modeling). It simplifies the traditionally complex manual design process, generating multiple energy-efficient design proposals within approximately 30 seconds. Users can select and modify specific building materials and equipment while accessing detailed product information. Integration with BIM enables real-time updates to design changes, improving efficiency for designers and clients. The service currently supports energy efficiency evaluations for non-residential buildings, with widespread adoption anticipated as energy efficiency standards become mandatory in 2025. Additionally, efforts are underway to expand the technology to cover domestic and international buildings, aiming for energy savings across entire urban areas rather than individual structures.
Creator: TOYODA Suzu, MORIKAMI Hirona, TOYODA Shoichiro, ISHIDA Takato
PM: HIRANO Yutaka
Development of Customizable Crop Sorting System for Farmers
To improve labor shortages and inefficiencies in crop sorting processes, the AI sorting machine "YasAI" has been developed, enabling automated sorting optimized for individual farms. The system achieves 80-90% accuracy in grading (premium-standard-good) and requires fewer datasets compared to conventional systems while maintaining high precision. Designed for transportability via light trucks, it is offered as a rental service, reducing maintenance needs during off-seasons. Additionally, the system features a digitally barrier-free UI, making it accessible to elderly farmers and lowering adoption barriers. Demonstration experiments in Hiroshima Prefecture confirmed compatibility with multiple crop varieties, paving the way for broader deployment. This technology is expected to reduce agricultural labor costs, improve operational efficiency, expand sales channels for individual farmers, and contribute to the sustainability of Japanese agriculture.
Creators: MIZUTANI Koyu, HIRATA Yuya, HIGUCHI Yuki, SUGIYAMA Shiho, SUGIYAMA Luna
Weight-Sensing Platform for Various Environments
The "Weight-Sensing Platform" project aims to establish social infrastructure supporting efficiency, safety, and health through the development of high-precision, cost-effective sensing devices and cloud systems with our foam-material and inductive sensing technology. The sensors feature high shape adaptability, maintainability, a wide dynamic range, and high sampling rates, enabling large-scale data collection. Data is stored, visualized, and analyzed in the cloud, accommodating various applications such as inventory management and health monitoring. Outreach activities have facilitated adoption across industries including healthcare, robotics, and automotive components. This technology contributes to operational efficiency, accident prevention, and health maintenance/disease prevention, addressing societal challenges such as labor shortages and promoting preventative healthcare.
Creators: MINAGAWA Tatsuya, NAGASE Takuya, TAGAMI Daichi
PM: FUJII Akihito
Development of an Knowledge Management System to Promote Proactive Value Creation in Organizations
"Peblo" is a next-generation knowledge management tool designed to facilitate proactive information sharing within organizations. Unlike traditional knowledge management systems that require formal documentation, Peblo enables users to effortlessly post and share everyday insights and ideas. AI-driven features like automatic tagging and content recommendations simplify the organization and utilization of accumulated information. The platform also supports active communication through reaction and collaborative editing functions. Integration with Slack is planned, allowing seamless adoption for teams using existing tools. Currently undergoing beta testing, Peblo is being refined for broader adoption in domestic and international markets. It is expected to be utilized across diverse fields, including businesses, research institutions, and educational environments, serving as a foundation for fostering innovation and collaboration.
Creators: NOZAKI Tomohiro, MIHASHI Yuki
Development of underwear device for measuring menstrual blood aimed at alleviating stress during menstruation
"Unleak" is a shorts-style menstrual blood detection device designed to reduce menstrual leakage risks and improve women's quality of life (QOL). The device uses optical sensors installed within the shorts to monitor the distribution of menstrual blood on sanitary napkins in real time, notifying users of replacement timing through vibration. Its non-contact design minimizes skin burden while allowing compatibility with existing sanitary products. User feedback from interviews and tests has been positive, prompting efforts toward miniaturization and feature enhancements. Moving forward, plans include commercialization and expansion into healthcare and sports sectors, with aims to broaden the application scope of its technology. Additionally, this technology holds potential for use in caregiving and infant diapers, offering solutions to various social challenges.
Creators: TSUNETSUGU Mai, YOSHIDA Saya, TAKAYASU Yuta, NOZAKI Naoki
PM: MIKI Hirofumi
AI-Driven Video System to Enhance Sports Coaching and Viewing Experiences
"Playbox" is an AI-driven video system that automates the capture, editing, and sharing of sports footage, designed for easy adoption by amateur and youth-level clubs with its low-cost setup. Utilizing commercial action cameras, the system automatically analyzes match footage to generate highlights and personalized videos for individual players. Equipped with cooling mechanisms and stitching technology, the camera units can cover large fields with minimal equipment. The system enhances coaching by providing efficient video analysis and improves viewing experiences by delivering personalized content to fans and families. Beta testing is underway with domestic clubs, and future plans include expansion into other sports and non-sports applications.
Creators: SCOTT Atom, UCHIDA Ikuma, SHRESTHA Pragyan
Development of HSI library and forest fire prevention system
"Development of HSI Library and Forest Fire Prevention System" harnesses hyperspectral imaging (HSI) technology to enable forest fire prediction and early warning. Compared to conventional visible light and infrared sensors, HSI offers higher precision in detecting vegetation changes and abnormalities. This project has developed an HSI data analysis library and AI-powered analysis engine, utilizing drones and satellites to collect extensive data. The technology is applicable not only for forest fire prevention but also for agriculture, environmental monitoring, and other fields. Plans include offering contract-based analysis services and launching a cloud-based subscription service. The initiative is expected to yield broad benefits, including environmental protection, reduced disaster management costs, and the creation of new markets in related industries.
Creators: TAKEDA Yuya、LEE Jaewon、TAKAHASHI Junichiro
Development of an AI-Powered Integrated Recruitment System
CaseMatch (https://casematch.jp/) is an AI-driven interview platform designed to streamline and standardize professional recruitment processes. Traditional interviews often face challenges in unifying evaluation criteria and implementing practical selection methods, which CaseMatch addresses through AI-generated case questions and response analysis. Users engage with weekly AI Interviews, honing their skills through AI interactions while receiving selection invitations from companies. Meanwhile, businesses benefit from utilizing a talent pool for efficient recruitment. The prototype stage garnered over 10,000 users, and the official version has been highly rated for its novelty in targeting specialized professions. Future plans include expanding into mid-career recruitment markets and other industries, promoting the adoption of AI interviews and enhancing their reliability.
Creators: YAMAMURO Kohei, MATSUMOTO Hiroki, NARA Ryoya, OYAMA Keigo, NODA Shiho, MIYAZAWA Sora
Change log
-
Dec 19, 2025
Released this page