Learning by demonstration for motion planning of upper-limb exoskeletons

Clemente Lauretti, Francesca Cordella, Anna Lisa Ciancio, Emilio Trigili, Jose Maria Catalan, Francisco Javier Badesa, Simona Crea, Silvio Marcello Pagliara, Silvia Sterzi, Nicola Vitiello, Nicolas Garcia Aracil, Loredana Zollo

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Copyright: © 2018 Lauretti, Cordella, Ciancio, Trigili, Catalan, Badesa, Crea, Pagliara, Sterzi, Vitiello, Garcia Aracil and Zollo. The reference joint position of upper-limb exoskeletons is typically obtained by means of Cartesian motion planners and inverse kinematics algorithms with the inverse Jacobian; this approach allows exploiting the available Degrees of Freedom (i.e. DoFs) of the robot kinematic chain to achieve the desired end-effector pose; however, if used to operate non-redundant exoskeletons, it does not ensure that anthropomorphic criteria are satisfied in the whole human-robot workspace. This paper proposes a motion planning system, based on Learning by Demonstration, for upper-limb exoskeletons that allow successfully assisting patients during Activities of Daily Living (ADLs) in unstructured environment, while ensuring that anthropomorphic criteria are satisfied in the whole human-robot workspace. The motion planning system combines Learning by Demonstration with the computation of Dynamic Motion Primitives and machine learning techniques to construct task- and patient-specific joint trajectories based on the learnt trajectories. System validation was carried out in simulation and in a real setting with a 4-DoF upper-limb exoskeleton, a 5-DoF wrist-hand exoskeleton and four patients with Limb Girdle Muscular Dystrophy. Validation was addressed to (i) compare the performance of the proposed motion planning with traditional methods; (ii) assess the generalization capabilities of the proposed method with respect to the environment variability. Three ADLs were chosen to validate the system: drinking, pouring and lifting a light sphere. The achieved results showed a 100% success rate in the task fulfillment, with a high level of generalization with respect to the environment variability. Moreover, an anthropomorphic configuration of the exoskeleton is always ensured.
Original languageEnglish
JournalFrontiers in Neurorobotics
Volume12
Issue numberFEB
DOIs
Publication statusPublished - Jan 1 2018

Fingerprint

Motion planning
Demonstrations
Robots
Trajectories
Inverse kinematics
End effectors
Learning systems
Kinematics
Exoskeleton (Robotics)

Keywords

  • Assistive robotics
  • Dynamics movement primitives
  • Learning by demonstration
  • Machine learning
  • Motion planning

Cite this

Lauretti, C., Cordella, F., Ciancio, A. L., Trigili, E., Catalan, J. M., Badesa, F. J., ... Zollo, L. (2018). Learning by demonstration for motion planning of upper-limb exoskeletons. Frontiers in Neurorobotics, 12(FEB). https://doi.org/10.3389/fnbot.2018.00005

Learning by demonstration for motion planning of upper-limb exoskeletons. / Lauretti, Clemente; Cordella, Francesca; Ciancio, Anna Lisa; Trigili, Emilio; Catalan, Jose Maria; Badesa, Francisco Javier; Crea, Simona; Pagliara, Silvio Marcello; Sterzi, Silvia; Vitiello, Nicola; Aracil, Nicolas Garcia; Zollo, Loredana.

In: Frontiers in Neurorobotics, Vol. 12, No. FEB, 01.01.2018.

Research output: Contribution to journalArticle

Lauretti, C, Cordella, F, Ciancio, AL, Trigili, E, Catalan, JM, Badesa, FJ, Crea, S, Pagliara, SM, Sterzi, S, Vitiello, N, Aracil, NG & Zollo, L 2018, 'Learning by demonstration for motion planning of upper-limb exoskeletons', Frontiers in Neurorobotics, vol. 12, no. FEB. https://doi.org/10.3389/fnbot.2018.00005
Lauretti C, Cordella F, Ciancio AL, Trigili E, Catalan JM, Badesa FJ et al. Learning by demonstration for motion planning of upper-limb exoskeletons. Frontiers in Neurorobotics. 2018 Jan 1;12(FEB). https://doi.org/10.3389/fnbot.2018.00005
Lauretti, Clemente ; Cordella, Francesca ; Ciancio, Anna Lisa ; Trigili, Emilio ; Catalan, Jose Maria ; Badesa, Francisco Javier ; Crea, Simona ; Pagliara, Silvio Marcello ; Sterzi, Silvia ; Vitiello, Nicola ; Aracil, Nicolas Garcia ; Zollo, Loredana. / Learning by demonstration for motion planning of upper-limb exoskeletons. In: Frontiers in Neurorobotics. 2018 ; Vol. 12, No. FEB.
@article{3018c5633e45405d888a9cf1e2466ad0,
title = "Learning by demonstration for motion planning of upper-limb exoskeletons",
abstract = "Copyright: {\circledC} 2018 Lauretti, Cordella, Ciancio, Trigili, Catalan, Badesa, Crea, Pagliara, Sterzi, Vitiello, Garcia Aracil and Zollo. The reference joint position of upper-limb exoskeletons is typically obtained by means of Cartesian motion planners and inverse kinematics algorithms with the inverse Jacobian; this approach allows exploiting the available Degrees of Freedom (i.e. DoFs) of the robot kinematic chain to achieve the desired end-effector pose; however, if used to operate non-redundant exoskeletons, it does not ensure that anthropomorphic criteria are satisfied in the whole human-robot workspace. This paper proposes a motion planning system, based on Learning by Demonstration, for upper-limb exoskeletons that allow successfully assisting patients during Activities of Daily Living (ADLs) in unstructured environment, while ensuring that anthropomorphic criteria are satisfied in the whole human-robot workspace. The motion planning system combines Learning by Demonstration with the computation of Dynamic Motion Primitives and machine learning techniques to construct task- and patient-specific joint trajectories based on the learnt trajectories. System validation was carried out in simulation and in a real setting with a 4-DoF upper-limb exoskeleton, a 5-DoF wrist-hand exoskeleton and four patients with Limb Girdle Muscular Dystrophy. Validation was addressed to (i) compare the performance of the proposed motion planning with traditional methods; (ii) assess the generalization capabilities of the proposed method with respect to the environment variability. Three ADLs were chosen to validate the system: drinking, pouring and lifting a light sphere. The achieved results showed a 100{\%} success rate in the task fulfillment, with a high level of generalization with respect to the environment variability. Moreover, an anthropomorphic configuration of the exoskeleton is always ensured.",
keywords = "Assistive robotics, Dynamics movement primitives, Learning by demonstration, Machine learning, Motion planning",
author = "Clemente Lauretti and Francesca Cordella and Ciancio, {Anna Lisa} and Emilio Trigili and Catalan, {Jose Maria} and Badesa, {Francisco Javier} and Simona Crea and Pagliara, {Silvio Marcello} and Silvia Sterzi and Nicola Vitiello and Aracil, {Nicolas Garcia} and Loredana Zollo",
year = "2018",
month = "1",
day = "1",
doi = "10.3389/fnbot.2018.00005",
language = "English",
volume = "12",
journal = "Frontiers in Neurorobotics",
issn = "1662-5218",
publisher = "Frontiers Research Foundation",
number = "FEB",

}

TY - JOUR

T1 - Learning by demonstration for motion planning of upper-limb exoskeletons

AU - Lauretti, Clemente

AU - Cordella, Francesca

AU - Ciancio, Anna Lisa

AU - Trigili, Emilio

AU - Catalan, Jose Maria

AU - Badesa, Francisco Javier

AU - Crea, Simona

AU - Pagliara, Silvio Marcello

AU - Sterzi, Silvia

AU - Vitiello, Nicola

AU - Aracil, Nicolas Garcia

AU - Zollo, Loredana

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Copyright: © 2018 Lauretti, Cordella, Ciancio, Trigili, Catalan, Badesa, Crea, Pagliara, Sterzi, Vitiello, Garcia Aracil and Zollo. The reference joint position of upper-limb exoskeletons is typically obtained by means of Cartesian motion planners and inverse kinematics algorithms with the inverse Jacobian; this approach allows exploiting the available Degrees of Freedom (i.e. DoFs) of the robot kinematic chain to achieve the desired end-effector pose; however, if used to operate non-redundant exoskeletons, it does not ensure that anthropomorphic criteria are satisfied in the whole human-robot workspace. This paper proposes a motion planning system, based on Learning by Demonstration, for upper-limb exoskeletons that allow successfully assisting patients during Activities of Daily Living (ADLs) in unstructured environment, while ensuring that anthropomorphic criteria are satisfied in the whole human-robot workspace. The motion planning system combines Learning by Demonstration with the computation of Dynamic Motion Primitives and machine learning techniques to construct task- and patient-specific joint trajectories based on the learnt trajectories. System validation was carried out in simulation and in a real setting with a 4-DoF upper-limb exoskeleton, a 5-DoF wrist-hand exoskeleton and four patients with Limb Girdle Muscular Dystrophy. Validation was addressed to (i) compare the performance of the proposed motion planning with traditional methods; (ii) assess the generalization capabilities of the proposed method with respect to the environment variability. Three ADLs were chosen to validate the system: drinking, pouring and lifting a light sphere. The achieved results showed a 100% success rate in the task fulfillment, with a high level of generalization with respect to the environment variability. Moreover, an anthropomorphic configuration of the exoskeleton is always ensured.

AB - Copyright: © 2018 Lauretti, Cordella, Ciancio, Trigili, Catalan, Badesa, Crea, Pagliara, Sterzi, Vitiello, Garcia Aracil and Zollo. The reference joint position of upper-limb exoskeletons is typically obtained by means of Cartesian motion planners and inverse kinematics algorithms with the inverse Jacobian; this approach allows exploiting the available Degrees of Freedom (i.e. DoFs) of the robot kinematic chain to achieve the desired end-effector pose; however, if used to operate non-redundant exoskeletons, it does not ensure that anthropomorphic criteria are satisfied in the whole human-robot workspace. This paper proposes a motion planning system, based on Learning by Demonstration, for upper-limb exoskeletons that allow successfully assisting patients during Activities of Daily Living (ADLs) in unstructured environment, while ensuring that anthropomorphic criteria are satisfied in the whole human-robot workspace. The motion planning system combines Learning by Demonstration with the computation of Dynamic Motion Primitives and machine learning techniques to construct task- and patient-specific joint trajectories based on the learnt trajectories. System validation was carried out in simulation and in a real setting with a 4-DoF upper-limb exoskeleton, a 5-DoF wrist-hand exoskeleton and four patients with Limb Girdle Muscular Dystrophy. Validation was addressed to (i) compare the performance of the proposed motion planning with traditional methods; (ii) assess the generalization capabilities of the proposed method with respect to the environment variability. Three ADLs were chosen to validate the system: drinking, pouring and lifting a light sphere. The achieved results showed a 100% success rate in the task fulfillment, with a high level of generalization with respect to the environment variability. Moreover, an anthropomorphic configuration of the exoskeleton is always ensured.

KW - Assistive robotics

KW - Dynamics movement primitives

KW - Learning by demonstration

KW - Machine learning

KW - Motion planning

UR - http://www.scopus.com/inward/record.url?scp=85042454067&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85042454067&partnerID=8YFLogxK

U2 - 10.3389/fnbot.2018.00005

DO - 10.3389/fnbot.2018.00005

M3 - Article

AN - SCOPUS:85042454067

VL - 12

JO - Frontiers in Neurorobotics

JF - Frontiers in Neurorobotics

SN - 1662-5218

IS - FEB

ER -