Research
Articles on us
Professor interview by Dept. of AIS (in Japanese)
An article by the public relations office of the Department of Advanced Interdisciplinary Studies. It briefly introduces the ideas of our study such as data-driven health management systems for systems like artificial satellites.
Lab description by Dept. of AIS
Our research topics and environment are introduced.
Research topics
Our research topics are not limited to those introduced here; we study broadly on machine learning and artificial intelligence.
Data-driven system health management
We use machine learning for tasks of system health management, such as anomaly detection and remaining useful life estimation. The developed methods have been applied to artificial satellites, industrial plants, and vehicles, etc.
Methods and application of probabilistic inference
Inference with uncertainty quantification is a key to principled decision making. We study not only general methodologies but also application in aerospace domains.
Planning, control, and reinforcement learning for autonomous robots
Path planning and control are fundamental building blocks of autonomous robots. We are interested in developing methods for planning and control, as well as in applying reinforcement learning to real mobile robots.
Synergy of machine learning and scientific models
Combining machine learning and mathematical models in scientific domains may create robust, partly interpretable, and flexible models. We study methods to achieve such combination.
Learning and analysis of dynamical systems
We are interested in learning and analysis of dynamical systems. Our targets range from classical state-space models to operator-theoretic perspectives.
Asteroid shape estimation & visual navigation
Estimating the shape of a target asteroid is an important step in asteroid exploration. We have been developing methods to estimate asteroid’s shape and spacecraft’s pose from monocular images.