Research Support Office Research Advancement Division. Tokyo University of Agriculture and Technology
| TEL | +81-42-367-5944 |
|---|---|
| FAX | +81-42-367-5946 |
This program is supported by MEXT’s scientific technology human resource development fee grant, "Program to Disseminate Tenure Tracking System".
Home > Introduction of our tenure-track faculties > Ariizumi Ryo
Ariizumi Ryo
| Affiliation | Institute of Engineering |
|---|---|
| Division | Division of Advanced Mechanical Systems Engineering |
| Research field | Robotics, Control Engineering |
| Keyword(S) | Bio-inspired robots, Data-driven control, Reinforcement learning |
| Url | https://web.tuat.ac.jp/~ariizumi-lab/index.html |
| Research experience | ･ Apr. 2014-Mar. 2015: JSPS Research Fellow (DC2) |
|---|---|
| Educational background | ･ Mar. 2010: B.S. (Eng.) Faculty of Engineering, Kyoto University |
| Awards | * The latest information is shown at the member's website. |
| Selected papers and publications | * The latest information is shown at the member's website. |
Bio-inspired robots, such as snake robots and legged robots, are examples of hyper-redundant systems, i.e., systems having many extra degrees of freedom (DOFs). The large number of DOFs enables the robots to perform many different tasks but also brings some difficulties in control. Because of this, many studies extract some features from the motion of biological counterparts or rely on reinforcement learning (RL) techniques. However, there are several problems. For example, biologically optimal motion is not necessarily suitable for robots. Another example is that it is hard to incorporate the prior knowledge obtained from control engineering or mechanics into RL, which prevents efficient learning.
In my research, along with the studies to deepen the understanding of the robots based on mechanics and control theory, I will try combining those methodologies to propose a new framework to generate intelligent motion of robots. Specifically, I will propose a new motion for snake robots that cannot be seen in nature to extend their capability. I will also seek a framework to incorporate the prior knowledge obtained from mechanics into RL, which can speed up the learning process. To this end, some techniques in control engineering will be an essential factor. Furthermore, I will tackle the problem of the automatic design of robots: i.e., the simultaneous optimization of the robot structure and its controller.
The Tenure Track program of the Tokyo University of Agriculture and Technology provides good support not only in finance but also in paper works. Thanks to such support, it is possible to start my research without taking much time on non-essential tasks related to the launch of my lab. I believe this environment is one of the bests in Japanese universities for young researchers.
Research on robots requires knowledge from many different fields. As the TUAT gives a good environment to make collaborations with other professors, it is a good place to perform such a multidisciplinary study. I will do my best in both research and education by utilizing the good environment provided by the university.
