Estimating perturbations from experience using neural networks and information transfer

Erik Berger, David Vogt, Steve Grehl, Bernhard Jung, Hani Ben Amor

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

In order to ensure safe operation, robots must be able to reliably detect behavior perturbations that result from unexpected physical interactions with their environment and human co-workers. While some robots provide firmware force sensors that generate rough force estimates, more accurate force measurements are usually achieved with dedicated force-torque sensors. However, such sensors are often heavy, expensive and require an additional power supply. In the case of lightweight manipulators, the already limited payload capabilities may be reduced in a significant way. This paper presents an experience-based approach for accurately estimating external forces being applied to a robot without the need for a forcetorque sensor. Using Information Transfer, a subset of sensors relevant to the executed behavior are identified from a larger set of internal sensors. Models mapping robot sensor data to force-torque measurements are learned using a neural network. These models can be used to predict the magnitude and direction of perturbations from affordable, proprioceptive sensors only. Experiments with a UR5 robot show that our method yields force estimates with accuracy comparable to a dedicated force-torque sensor. Moreover, our method yields a substantial improvement in accuracy over force-torque values provided by the robot firmware.

Original languageEnglish (US)
Title of host publicationIROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages176-181
Number of pages6
Volume2016-November
ISBN (Electronic)9781509037629
DOIs
StatePublished - Nov 28 2016
Event2016 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2016 - Daejeon, Korea, Republic of
Duration: Oct 9 2016Oct 14 2016

Other

Other2016 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2016
CountryKorea, Republic of
CityDaejeon
Period10/9/1610/14/16

Fingerprint

Neural networks
Sensors
Robots
Firmware
Torque
Force measurement
Torque measurement
Manipulators
Experiments

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Software
  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

Berger, E., Vogt, D., Grehl, S., Jung, B., & Ben Amor, H. (2016). Estimating perturbations from experience using neural networks and information transfer. In IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (Vol. 2016-November, pp. 176-181). [7759052] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IROS.2016.7759052

Estimating perturbations from experience using neural networks and information transfer. / Berger, Erik; Vogt, David; Grehl, Steve; Jung, Bernhard; Ben Amor, Hani.

IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems. Vol. 2016-November Institute of Electrical and Electronics Engineers Inc., 2016. p. 176-181 7759052.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Berger, E, Vogt, D, Grehl, S, Jung, B & Ben Amor, H 2016, Estimating perturbations from experience using neural networks and information transfer. in IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems. vol. 2016-November, 7759052, Institute of Electrical and Electronics Engineers Inc., pp. 176-181, 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2016, Daejeon, Korea, Republic of, 10/9/16. https://doi.org/10.1109/IROS.2016.7759052
Berger E, Vogt D, Grehl S, Jung B, Ben Amor H. Estimating perturbations from experience using neural networks and information transfer. In IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems. Vol. 2016-November. Institute of Electrical and Electronics Engineers Inc. 2016. p. 176-181. 7759052 https://doi.org/10.1109/IROS.2016.7759052
Berger, Erik ; Vogt, David ; Grehl, Steve ; Jung, Bernhard ; Ben Amor, Hani. / Estimating perturbations from experience using neural networks and information transfer. IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems. Vol. 2016-November Institute of Electrical and Electronics Engineers Inc., 2016. pp. 176-181
@inproceedings{824ef83f701f44919b6f569e703ba13f,
title = "Estimating perturbations from experience using neural networks and information transfer",
abstract = "In order to ensure safe operation, robots must be able to reliably detect behavior perturbations that result from unexpected physical interactions with their environment and human co-workers. While some robots provide firmware force sensors that generate rough force estimates, more accurate force measurements are usually achieved with dedicated force-torque sensors. However, such sensors are often heavy, expensive and require an additional power supply. In the case of lightweight manipulators, the already limited payload capabilities may be reduced in a significant way. This paper presents an experience-based approach for accurately estimating external forces being applied to a robot without the need for a forcetorque sensor. Using Information Transfer, a subset of sensors relevant to the executed behavior are identified from a larger set of internal sensors. Models mapping robot sensor data to force-torque measurements are learned using a neural network. These models can be used to predict the magnitude and direction of perturbations from affordable, proprioceptive sensors only. Experiments with a UR5 robot show that our method yields force estimates with accuracy comparable to a dedicated force-torque sensor. Moreover, our method yields a substantial improvement in accuracy over force-torque values provided by the robot firmware.",
author = "Erik Berger and David Vogt and Steve Grehl and Bernhard Jung and {Ben Amor}, Hani",
year = "2016",
month = "11",
day = "28",
doi = "10.1109/IROS.2016.7759052",
language = "English (US)",
volume = "2016-November",
pages = "176--181",
booktitle = "IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

TY - GEN

T1 - Estimating perturbations from experience using neural networks and information transfer

AU - Berger, Erik

AU - Vogt, David

AU - Grehl, Steve

AU - Jung, Bernhard

AU - Ben Amor, Hani

PY - 2016/11/28

Y1 - 2016/11/28

N2 - In order to ensure safe operation, robots must be able to reliably detect behavior perturbations that result from unexpected physical interactions with their environment and human co-workers. While some robots provide firmware force sensors that generate rough force estimates, more accurate force measurements are usually achieved with dedicated force-torque sensors. However, such sensors are often heavy, expensive and require an additional power supply. In the case of lightweight manipulators, the already limited payload capabilities may be reduced in a significant way. This paper presents an experience-based approach for accurately estimating external forces being applied to a robot without the need for a forcetorque sensor. Using Information Transfer, a subset of sensors relevant to the executed behavior are identified from a larger set of internal sensors. Models mapping robot sensor data to force-torque measurements are learned using a neural network. These models can be used to predict the magnitude and direction of perturbations from affordable, proprioceptive sensors only. Experiments with a UR5 robot show that our method yields force estimates with accuracy comparable to a dedicated force-torque sensor. Moreover, our method yields a substantial improvement in accuracy over force-torque values provided by the robot firmware.

AB - In order to ensure safe operation, robots must be able to reliably detect behavior perturbations that result from unexpected physical interactions with their environment and human co-workers. While some robots provide firmware force sensors that generate rough force estimates, more accurate force measurements are usually achieved with dedicated force-torque sensors. However, such sensors are often heavy, expensive and require an additional power supply. In the case of lightweight manipulators, the already limited payload capabilities may be reduced in a significant way. This paper presents an experience-based approach for accurately estimating external forces being applied to a robot without the need for a forcetorque sensor. Using Information Transfer, a subset of sensors relevant to the executed behavior are identified from a larger set of internal sensors. Models mapping robot sensor data to force-torque measurements are learned using a neural network. These models can be used to predict the magnitude and direction of perturbations from affordable, proprioceptive sensors only. Experiments with a UR5 robot show that our method yields force estimates with accuracy comparable to a dedicated force-torque sensor. Moreover, our method yields a substantial improvement in accuracy over force-torque values provided by the robot firmware.

UR - http://www.scopus.com/inward/record.url?scp=85006401324&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85006401324&partnerID=8YFLogxK

U2 - 10.1109/IROS.2016.7759052

DO - 10.1109/IROS.2016.7759052

M3 - Conference contribution

VL - 2016-November

SP - 176

EP - 181

BT - IROS 2016 - 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems

PB - Institute of Electrical and Electronics Engineers Inc.

ER -