A study on the algorithms for accelerating human motor learning under human-robot cooperative motor learning system

2022 Fiscal Year Final Research Report

• Project/Area Number: 20H02111
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 20020:Robotics and intelligent system-related
• Research Institution: Tokyo University of Agriculture and Technology
• Principal Investigator: Kondo Toshiyuki 東京農工大学, 工学(系)研究科(研究院), 教授 (60323820)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Keywords: 協調運動学習 / 動的機能結合解析 / リハビリ

Outline of Final Research Achievements

In this study, we developed four robotic agents, which could manipulate cooperative motor learning, and compared the effect of their intervening strategies (i.e., novice, expert, level-up, and skill-level-adjustment agents) on individual motor performance. The result suggests that the novice agent realizes significantly higher posterior motor performance than the expert agent. Moreover, we found that skill-level adjust agent, which grows with the subject's motor skill level, is able to provide even better motor support than the others. We also proposed a dynamic functional connectivity analysis method that combines tensor decomposition and TVGL methods. We applied the method to sleep EEG, --- the brain states are medically labeled ---, and verified the validity.

Free Research Field

知能情報学

Academic Significance and Societal Importance of the Research Achievements

本研究では、ヒトとロボットの協調運動学習を対象に、ロボットエージェントの制御アルゴリズムを4種類（初心者、熟練者、レベル調節、固定成長）考案し、被験者実験を行った。実験の結果、被験者の運動技能レベルにあわせてともに成長するレベル調節エージェントが最良であることが示された。ロボットリハビリテーションにおける最適な支援の量は患者ごとに異なり、また患者の機能改善とともに変化すると考えられることから、本研究の成果は、将来のロボットリハビリテーション開発にとって有益と考えられる。