anticache

ICS Research Seminar on "Learning Assistive Strategies for Exoskeleton Robots from User-Robot Physical Interaction"

01.03.2017


ICS Research Seminar on: "Learning Assistive Strategies for Exoskeleton Robots from User-Robot Physical Interaction"

Presenters: Prof. Tsukasa Ogasawara, Assoc. Prof. Takamitsu Masubara

Talk 1 (10min) Title: Introduction of NAIST IS

Presenter: Prof. Tsukasa Ogasawara, Dean of the Graduate School of Information Science, Nara Institute of Science and Technology, Nara, Japan

Talk 2 (30min) Title: Learning Assistive Strategies for Exoskeleton Robots from User-Robot Physical Interaction

Presenter: Assoc.Prof. Takamitsu Matsubara

Abstract: Social demand for exoskeleton robots that physically assist humans has been increasing in various situations due to the demographic trends of aging populations. In exoskeleton robots, an assistive strategy is a key ingredient. Since interactions between users and exoskeleton robots are bidirectional, the assistive strategy design problem is complex and challenging. We explore a data-driven learning approach for designing assistive strategies for exoskeletons from user-robot physical interaction. We formulate the learning problem of assistive strategies as a policy search problem and exploit a data-effcient model-based reinforcement learning framework. Instead of explicitly providing the desired trajectories in the cost function, our cost function only considers the user's muscular effort measured by @electromyography signals (EMGs) to learn the assistive strategies. The key assumption behind is that the user is instructed to perform the task by his/her own intended movements. Since the EMGs are observed when the intended movements are achieved by the user own muscle efforts rather than the robot's assistance, EMGs can be interpreted as the gcost h of the current assistance.

3. March 2017 at 10:00 h ICS Karlstraße 45, 2. Floor, Seminar room 2026