Estimation of perturbations in robotic behavior using dynamic mode decomposition

Erik Berger, Mark Sastuba, David Vogt, Bernhard Jung, Hani Ben Amor

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

Physical human-robot interaction tasks require robots that can detect and react to external perturbations caused by the human partner. In this contribution, we present a machine learning approach for detecting, estimating, and compensating for such external perturbations using only input from standard sensors. This machine learning approach makes use of Dynamic Mode Decomposition (DMD), a data processing technique developed in the field of fluid dynamics, which is applied to robotics for the first time. DMD is able to isolate the dynamics of a nonlinear system and is therefore well suited for separating noise from regular oscillations in sensor readings during cyclic robot movements. In a training phase, a DMD model for behavior-specific parameter configurations is learned. During task execution, the robot must estimate the external forces exerted by a human interaction partner. We compare the DMD-based approach to other interpolation schemes. A variant, sparsity promoting DMD, is particularly well suited for high-noise sensors. Results of a user study show that our DMD-based machine learning approach can be used to design physical human-robot interaction techniques that not only result in robust robot behavior but also enjoy a high usability.

Original languageEnglish (US)
Pages (from-to)331-343
Number of pages13
JournalAdvanced Robotics
Volume29
Issue number5
DOIs
StatePublished - Mar 4 2015
Externally publishedYes

Fingerprint

Robotics
Decomposition
Robots
Learning systems
Human robot interaction
Sensors
Fluid dynamics
Nonlinear systems
Interpolation

Keywords

  • dynamic mode decomposition
  • external perturbation
  • model learning
  • physical humanrobot interaction
  • usability in humanrobot interaction

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Human-Computer Interaction
  • Computer Science Applications
  • Hardware and Architecture
  • Software

Cite this

Estimation of perturbations in robotic behavior using dynamic mode decomposition. / Berger, Erik; Sastuba, Mark; Vogt, David; Jung, Bernhard; Ben Amor, Hani.

In: Advanced Robotics, Vol. 29, No. 5, 04.03.2015, p. 331-343.

Research output: Contribution to journalArticle

Berger, Erik ; Sastuba, Mark ; Vogt, David ; Jung, Bernhard ; Ben Amor, Hani. / Estimation of perturbations in robotic behavior using dynamic mode decomposition. In: Advanced Robotics. 2015 ; Vol. 29, No. 5. pp. 331-343.
@article{640bcf27f232449e9fd65a992ec03a3f,
title = "Estimation of perturbations in robotic behavior using dynamic mode decomposition",
abstract = "Physical human-robot interaction tasks require robots that can detect and react to external perturbations caused by the human partner. In this contribution, we present a machine learning approach for detecting, estimating, and compensating for such external perturbations using only input from standard sensors. This machine learning approach makes use of Dynamic Mode Decomposition (DMD), a data processing technique developed in the field of fluid dynamics, which is applied to robotics for the first time. DMD is able to isolate the dynamics of a nonlinear system and is therefore well suited for separating noise from regular oscillations in sensor readings during cyclic robot movements. In a training phase, a DMD model for behavior-specific parameter configurations is learned. During task execution, the robot must estimate the external forces exerted by a human interaction partner. We compare the DMD-based approach to other interpolation schemes. A variant, sparsity promoting DMD, is particularly well suited for high-noise sensors. Results of a user study show that our DMD-based machine learning approach can be used to design physical human-robot interaction techniques that not only result in robust robot behavior but also enjoy a high usability.",
keywords = "dynamic mode decomposition, external perturbation, model learning, physical humanrobot interaction, usability in humanrobot interaction",
author = "Erik Berger and Mark Sastuba and David Vogt and Bernhard Jung and {Ben Amor}, Hani",
year = "2015",
month = "3",
day = "4",
doi = "10.1080/01691864.2014.981292",
language = "English (US)",
volume = "29",
pages = "331--343",
journal = "Advanced Robotics",
issn = "0169-1864",
publisher = "Taylor and Francis Ltd.",
number = "5",

}

TY - JOUR

T1 - Estimation of perturbations in robotic behavior using dynamic mode decomposition

AU - Berger, Erik

AU - Sastuba, Mark

AU - Vogt, David

AU - Jung, Bernhard

AU - Ben Amor, Hani

PY - 2015/3/4

Y1 - 2015/3/4

N2 - Physical human-robot interaction tasks require robots that can detect and react to external perturbations caused by the human partner. In this contribution, we present a machine learning approach for detecting, estimating, and compensating for such external perturbations using only input from standard sensors. This machine learning approach makes use of Dynamic Mode Decomposition (DMD), a data processing technique developed in the field of fluid dynamics, which is applied to robotics for the first time. DMD is able to isolate the dynamics of a nonlinear system and is therefore well suited for separating noise from regular oscillations in sensor readings during cyclic robot movements. In a training phase, a DMD model for behavior-specific parameter configurations is learned. During task execution, the robot must estimate the external forces exerted by a human interaction partner. We compare the DMD-based approach to other interpolation schemes. A variant, sparsity promoting DMD, is particularly well suited for high-noise sensors. Results of a user study show that our DMD-based machine learning approach can be used to design physical human-robot interaction techniques that not only result in robust robot behavior but also enjoy a high usability.

AB - Physical human-robot interaction tasks require robots that can detect and react to external perturbations caused by the human partner. In this contribution, we present a machine learning approach for detecting, estimating, and compensating for such external perturbations using only input from standard sensors. This machine learning approach makes use of Dynamic Mode Decomposition (DMD), a data processing technique developed in the field of fluid dynamics, which is applied to robotics for the first time. DMD is able to isolate the dynamics of a nonlinear system and is therefore well suited for separating noise from regular oscillations in sensor readings during cyclic robot movements. In a training phase, a DMD model for behavior-specific parameter configurations is learned. During task execution, the robot must estimate the external forces exerted by a human interaction partner. We compare the DMD-based approach to other interpolation schemes. A variant, sparsity promoting DMD, is particularly well suited for high-noise sensors. Results of a user study show that our DMD-based machine learning approach can be used to design physical human-robot interaction techniques that not only result in robust robot behavior but also enjoy a high usability.

KW - dynamic mode decomposition

KW - external perturbation

KW - model learning

KW - physical humanrobot interaction

KW - usability in humanrobot interaction

UR - http://www.scopus.com/inward/record.url?scp=84925615527&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84925615527&partnerID=8YFLogxK

U2 - 10.1080/01691864.2014.981292

DO - 10.1080/01691864.2014.981292

M3 - Article

AN - SCOPUS:84925615527

VL - 29

SP - 331

EP - 343

JO - Advanced Robotics

JF - Advanced Robotics

SN - 0169-1864

IS - 5

ER -