Learning control in robot-assisted rehabilitation of motor skills – a review

Shou-Han Zhou, Justin Fong, Vincent Crocher, Ying Tan, Denny Oetomo, Iven Mareels

Research output: Contribution to journalReview ArticleResearchpeer-review

Abstract

The key idea in iterative learning control is captured by the intuition of ‘practice makes perfect’. The underlying learning is based on a gradient descent algorithm iteratively optimising an appropriate input–output measured criterion. How this paradigm is used to model quantitatively, at an input/output level, the learning that happens in the context of human motor skill learning is discussed in this note. Experimental studies of human motor learning, in robotically controlled environments, indicate that a model consisting of a classical (iterative) learning control augmented with an appropriate kinematic model of human motor motion fits the observed human learning behaviour well. In the context of the rehabilitation of motor skills, such models promise better human–machine interfaces that extend the capability and capacity of rehabilitation clinicians by creating effective robot–patient–clinician feedback loops. The economic promise of robot-assisted rehabilitation is to greatly extend the intervention capacity above what presently can be achieved by rehabilitation systems: addressing the needs of more people, over longer periods of time and at a distance in the comfort of their own personal environment. Moreover, the robot platforms provide for a more rigorous and quantitative evaluation of the patient’s motor skill across the entire personal rehabilitation trajectory, which opens up opportunities for improved, more individually tuned rehabilitation regimes.

Original languageEnglish
Pages (from-to)19-43
Number of pages25
JournalJournal of Control and Decision
Volume3
Issue number1
DOIs
Publication statusPublished - Jan 2016
Externally publishedYes

Keywords

  • human motor learning
  • learning control
  • motor adaptation
  • rehabilitation robots

Cite this

Zhou, Shou-Han ; Fong, Justin ; Crocher, Vincent ; Tan, Ying ; Oetomo, Denny ; Mareels, Iven. / Learning control in robot-assisted rehabilitation of motor skills – a review. In: Journal of Control and Decision. 2016 ; Vol. 3, No. 1. pp. 19-43.
@article{faa85fe025694e76bd4e804429aa0c20,
title = "Learning control in robot-assisted rehabilitation of motor skills – a review",
abstract = "The key idea in iterative learning control is captured by the intuition of ‘practice makes perfect’. The underlying learning is based on a gradient descent algorithm iteratively optimising an appropriate input–output measured criterion. How this paradigm is used to model quantitatively, at an input/output level, the learning that happens in the context of human motor skill learning is discussed in this note. Experimental studies of human motor learning, in robotically controlled environments, indicate that a model consisting of a classical (iterative) learning control augmented with an appropriate kinematic model of human motor motion fits the observed human learning behaviour well. In the context of the rehabilitation of motor skills, such models promise better human–machine interfaces that extend the capability and capacity of rehabilitation clinicians by creating effective robot–patient–clinician feedback loops. The economic promise of robot-assisted rehabilitation is to greatly extend the intervention capacity above what presently can be achieved by rehabilitation systems: addressing the needs of more people, over longer periods of time and at a distance in the comfort of their own personal environment. Moreover, the robot platforms provide for a more rigorous and quantitative evaluation of the patient’s motor skill across the entire personal rehabilitation trajectory, which opens up opportunities for improved, more individually tuned rehabilitation regimes.",
keywords = "human motor learning, learning control, motor adaptation, rehabilitation robots",
author = "Shou-Han Zhou and Justin Fong and Vincent Crocher and Ying Tan and Denny Oetomo and Iven Mareels",
year = "2016",
month = "1",
doi = "10.1080/23307706.2015.1129295",
language = "English",
volume = "3",
pages = "19--43",
journal = "Journal of Control and Decision",
issn = "2330-7706",
number = "1",

}

Learning control in robot-assisted rehabilitation of motor skills – a review. / Zhou, Shou-Han; Fong, Justin; Crocher, Vincent; Tan, Ying; Oetomo, Denny; Mareels, Iven.

In: Journal of Control and Decision, Vol. 3, No. 1, 01.2016, p. 19-43.

Research output: Contribution to journalReview ArticleResearchpeer-review

TY - JOUR

T1 - Learning control in robot-assisted rehabilitation of motor skills – a review

AU - Zhou, Shou-Han

AU - Fong, Justin

AU - Crocher, Vincent

AU - Tan, Ying

AU - Oetomo, Denny

AU - Mareels, Iven

PY - 2016/1

Y1 - 2016/1

N2 - The key idea in iterative learning control is captured by the intuition of ‘practice makes perfect’. The underlying learning is based on a gradient descent algorithm iteratively optimising an appropriate input–output measured criterion. How this paradigm is used to model quantitatively, at an input/output level, the learning that happens in the context of human motor skill learning is discussed in this note. Experimental studies of human motor learning, in robotically controlled environments, indicate that a model consisting of a classical (iterative) learning control augmented with an appropriate kinematic model of human motor motion fits the observed human learning behaviour well. In the context of the rehabilitation of motor skills, such models promise better human–machine interfaces that extend the capability and capacity of rehabilitation clinicians by creating effective robot–patient–clinician feedback loops. The economic promise of robot-assisted rehabilitation is to greatly extend the intervention capacity above what presently can be achieved by rehabilitation systems: addressing the needs of more people, over longer periods of time and at a distance in the comfort of their own personal environment. Moreover, the robot platforms provide for a more rigorous and quantitative evaluation of the patient’s motor skill across the entire personal rehabilitation trajectory, which opens up opportunities for improved, more individually tuned rehabilitation regimes.

AB - The key idea in iterative learning control is captured by the intuition of ‘practice makes perfect’. The underlying learning is based on a gradient descent algorithm iteratively optimising an appropriate input–output measured criterion. How this paradigm is used to model quantitatively, at an input/output level, the learning that happens in the context of human motor skill learning is discussed in this note. Experimental studies of human motor learning, in robotically controlled environments, indicate that a model consisting of a classical (iterative) learning control augmented with an appropriate kinematic model of human motor motion fits the observed human learning behaviour well. In the context of the rehabilitation of motor skills, such models promise better human–machine interfaces that extend the capability and capacity of rehabilitation clinicians by creating effective robot–patient–clinician feedback loops. The economic promise of robot-assisted rehabilitation is to greatly extend the intervention capacity above what presently can be achieved by rehabilitation systems: addressing the needs of more people, over longer periods of time and at a distance in the comfort of their own personal environment. Moreover, the robot platforms provide for a more rigorous and quantitative evaluation of the patient’s motor skill across the entire personal rehabilitation trajectory, which opens up opportunities for improved, more individually tuned rehabilitation regimes.

KW - human motor learning

KW - learning control

KW - motor adaptation

KW - rehabilitation robots

UR - http://www.scopus.com/inward/record.url?scp=85036456842&partnerID=8YFLogxK

U2 - 10.1080/23307706.2015.1129295

DO - 10.1080/23307706.2015.1129295

M3 - Review Article

VL - 3

SP - 19

EP - 43

JO - Journal of Control and Decision

JF - Journal of Control and Decision

SN - 2330-7706

IS - 1

ER -