The University of Aizu, in collaboration with Interstellar Technologies Inc. (*3), the lead implementing organization, will work on the development of a low-power control computation accelerator for real-time flight control of microsatellites as part of the commissioned project "Development of Formation Flight Technology Enabling Precise Control of a Large Number of Spacecraft" (*2), under the "High-Precision Satellite Formation Flight Technology" initiative (*1) of the Space Strategy Fund.
Senior Associate Professor TOMIOKA Yoichi of the University of Aizu has been engaged in energy-efficient AI acceleration research, including the development of the Soft Tensor Processor--an AI accelerator with high power efficiency--under a project commissioned by the New Energy and Industrial Technology Development Organization (NEDO), titled "Technology Development of AI Chips and Next-Generation Computing for High-Efficiency and High-Speed Processing."
Furthermore, Associate Professor Tomioka is also involved in the JST "PRESTO" program (*Presto: Precursory Research for Embryonic Science and Technology*), under the research area of "Trustworthy AI Infrastructure Technologies," where he is working on the realization of sustainable and high-efficiency AI systems. This research includes the development of reliable AI models and accelerators capable of maintaining high-precision inference processing even under conditions such as circuit aging and data corruption caused by radiation. This technology is expected to significantly advance the deployment of edge AI systems in space and extreme environments in the future.
Reference URL: https://u-aizu.ac.jp/~ytomioka/research/presto/presto.html
The commissioned project also includes participation from Professor DEMURA Hirohide, Director of the Space Information Science Center--an organization that has contributed to numerous space development projects, including the world's first asteroid sample return missions such as "Hayabusa" and "Hayabusa2." Additionally, from the IoT Cluster, which has been engaged in low-power edge AI R\&D, Professor SAITO Hiroshi (Cluster Leader), Senior Associate Professors Yoichi Tomioka and OKUYAMA Yuichi, and Assistant Professor SEMBA Shogo are also participating. Leveraging the University of Aizu's previously developed high-efficiency, fault-tolerant AI accelerator technologies, the project aims to develop a low-power and radiation-tolerant control computation accelerator for real-time flight control of microsatellites.
Through cutting-edge edge AI technology and innovative industry-academia collaboration, the University of Aizu continues to contribute to space development.
*1 Space Strategy Fund (Phase 1) - High-Precision Satellite Formation Flight Technology
*3 Interstellar Technologies Inc.