Robot Learning

Research Staff

• 

  Professor
  Takamitsu MATSUBARA

• 

  Associate Professor
  Kazuki SHIBATA

• 

  Affiliate Associate Professor
  Kenta HANADA

• 

  Assistant Professor
  Yoshihisa TSURUMINE

• 

  Assistant Professor
  Hikaru SASAKI

• 

  Assistant Professor
  Yuhwan KWON

• 

  Assistant Professor
  Cheng-yu KUO

• 

  Assistant Professor
  Yuki KADOKAWA

• 

  Affiliate Assistant Professor
  Hirotaka TAHARA

E-mail	takam-m@is.naist.jp shibata.kazuki@naist.ac.jp hanada@is.naist.jp tsurumine.yoshihisa@is.naist.jp sasaki.hikaru@is.naist.jp y-kwon@is.naist.jp chengyu.kuo@naist.ac.jp kadokawa.yuki@naist.ac.jp
To the site	https://sites.google.com/view/naist-robot-learning-en?authuser=0

Required knowledge and skills for starting research

Robot learning, the integration of machine learning and artificial intelligence with robotics, control engineering, signal processing, optimization, and mechatronics, offers the potential to develop unique approaches by leveraging one's strengths and experiences in areas such as robot competitions, programming contests, and crafting. We encourage you to take on the challenge of robot learning research.

Guidance philosophy of the laboratory

In the master's program, students set research topics in consultation with faculty, exploring solutions and conducting experiments with actual robots, following numerical simulations. Under faculty guidance, students aim to present their research internationally and publish papers in English. The doctoral program supports students in becoming independent researchers who can identify significant research topics and write papers in English by themselves.

Skills acquired through the research

In addition to gaining various knowledge about robot learning, you will acquire skills and sensibilities related to real-world robot learning systems construction, including working with actual robots and sensors, system integration, and data collection cost awareness. Through research activities, you will develop a full range of skills needed by engineers and researchers in society, including how to proceed with basic and applied research, summarize findings, and present them. We also nurture talent that can perform internationally through collaborations and internships with overseas research institutions. Participation in industry-academia joint research further cultivates practical skills. For beginners, we offer textbook reviews and reading groups on the latest papers.

We are advancing research in robot learning, an area that integrates machine learning and robotics, aimed at achieving advanced perception, judgment, and action functions in robots operating in human-centered environments. The core of this research includes probability, statistics, Bayesian theory, and deep learning, and involves designing learning algorithms for robots using control engineering, robotic engineering, and mathematical optimization. We are working on practical technologies for industrial/collaborative robots and assistive robots using cutting-edge robotic facilities (Nextage, UR5/3, Baxter, Unitree A1/Go1, OP3 Humanoid). We are also challenging the societal implementation of machine learning and artificial intelligence through industry-academia joint research.

• Reinforcement learning

• Imitation learning

• Human-robot collaboration
• Assistive robots
• Real-world applications