Learning and Robotic Imitation of Therapist’s Behavior for Rehabilitation Therapy

• Author / Creator

  Carlos Manuel Martínez

• The demand for rehabilitation services has increased in recent years due to population aging. Due to the limitations of therapists’ time and healthcare resources, robot-assisted therapy is becoming an appealing, powerful and economical solution. In this thesis, and in order to reach a long term goal, we propose different solutions that combine Learning from Demonstration (LfD) algorithms and robotic rehabilitation to save the therapist’s time and reduce the therapy costs as well as the patient’s recovery time. The target of this work is to show how medical robotics can be used in combination with LfD algorithms to learn and reproduce the therapist’s behavior during therapy based on ADL’s. In this thesis, three different tasks and experiments are presented. First, a telerehabilitation system to perform a unimanual cooperative task using LfD algorithms is presented. The second experiment targets ADLs that involve periodic motion in a 2D space. Later, a 2D reaching motion control task, as well as a 2D force control task are presented using a different LfD algorithm that helps to ensure the global asymptotic stability of the system. This thesis presents a step forward in the robotics rehabilitation context. By using LfD algorithms, we show that there is a new paradigm in the rehabilitation field where the robots can learn the therapist’s behavior and reproduce it even for complex tasks.

• Subjects / Keywords

  • Rehabilitation
  • machine learning
  • Robotics

• Graduation date

  Fall 2018

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R3610W78T

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.