Loading [MathJax]/extensions/tex2jax.js

Quantifying Uncertainties of Contact Classifications in a Human-Robot Collaboration with Parallel Robots

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Research Organisations

Details

Original languageEnglish
Title of host publicationHuman-Friendly Robotics 2023
Subtitle of host publicationHFR: 16th International Workshop on Human-Friendly Robotics
EditorsChristina Piazza, Patricia Capsi-Morales, Luis Figueredo, Manuel Keppler, Hinrich Schütze
Place of PublicationCham
Pages137-150
Number of pages14
ISBN (electronic)978-3-031-55002-7
Publication statusPublished - 10 Mar 2024
Event16th International Workshop on Human-Friendly Robotics (HFR 2023) - München, Germany
Duration: 20 Sept 202321 Sept 2023

Publication series

NameSpringer Proceedings in Advanced Robotics (SPAR)
Number29
ISSN (Print)2511-1256
ISSN (electronic)2511-1264

Abstract

In human-robot collaboration, unintentional physical contacts occur in the form of collisions and clamping, which must be detected and classified separately for a reaction. If certain collision or clamping situations are misclassified, reactions might occur that make the true contact case more dangerous. This work analyzes data-driven modeling based on physically modeled features like estimated external forces for clamping and collision classification with a real parallel robot. The prediction reliability of a feedforward neural network is investigated. Quantification of the classification uncertainty enables the distinction between safe versus unreliable classifications and optimal reactions like a retraction movement for collisions, structure opening for the clamping joint, and a fallback reaction in the form of a zero-g mode. This hypothesis is tested with experimental data of clamping and collision cases by analyzing dangerous misclassifications and then reducing them by the proposed uncertainty quantification. Finally, it is investigated how the approach of this work influences correctly classified clamping and collision scenarios.

Keywords

    cs.RO, cs.SY, eess.SY, parallel robots, human-robot collaboration, data-driven modeling

ASJC Scopus subject areas

Cite this

Quantifying Uncertainties of Contact Classifications in a Human-Robot Collaboration with Parallel Robots. / Mohammad, Aran; Muscheid, Hendrik; Schappler, Moritz et al.
Human-Friendly Robotics 2023: HFR: 16th International Workshop on Human-Friendly Robotics. ed. / Christina Piazza; Patricia Capsi-Morales; Luis Figueredo; Manuel Keppler; Hinrich Schütze. 1. ed. Cham, 2024. p. 137-150 (Springer Proceedings in Advanced Robotics (SPAR); No. 29).

Research output: Chapter in book/report/conference proceedingConference contributionResearchpeer review

Mohammad, A, Muscheid, H, Schappler, M & Seel, T 2024, Quantifying Uncertainties of Contact Classifications in a Human-Robot Collaboration with Parallel Robots. in C Piazza, P Capsi-Morales, L Figueredo, M Keppler & H Schütze (eds), Human-Friendly Robotics 2023: HFR: 16th International Workshop on Human-Friendly Robotics. 1. edn, Springer Proceedings in Advanced Robotics (SPAR), no. 29, Cham, pp. 137-150, 16th International Workshop on Human-Friendly Robotics (HFR 2023), München, Germany, 20 Sept 2023. https://doi.org/10.1007/978-3-031-55000-3_10
Mohammad, A., Muscheid, H., Schappler, M., & Seel, T. (2024). Quantifying Uncertainties of Contact Classifications in a Human-Robot Collaboration with Parallel Robots. In C. Piazza, P. Capsi-Morales, L. Figueredo, M. Keppler, & H. Schütze (Eds.), Human-Friendly Robotics 2023: HFR: 16th International Workshop on Human-Friendly Robotics (1. ed., pp. 137-150). (Springer Proceedings in Advanced Robotics (SPAR); No. 29).. https://doi.org/10.1007/978-3-031-55000-3_10
Mohammad A, Muscheid H, Schappler M, Seel T. Quantifying Uncertainties of Contact Classifications in a Human-Robot Collaboration with Parallel Robots. In Piazza C, Capsi-Morales P, Figueredo L, Keppler M, Schütze H, editors, Human-Friendly Robotics 2023: HFR: 16th International Workshop on Human-Friendly Robotics. 1. ed. Cham. 2024. p. 137-150. (Springer Proceedings in Advanced Robotics (SPAR); 29). doi: 10.1007/978-3-031-55000-3_10
Mohammad, Aran ; Muscheid, Hendrik ; Schappler, Moritz et al. / Quantifying Uncertainties of Contact Classifications in a Human-Robot Collaboration with Parallel Robots. Human-Friendly Robotics 2023: HFR: 16th International Workshop on Human-Friendly Robotics. editor / Christina Piazza ; Patricia Capsi-Morales ; Luis Figueredo ; Manuel Keppler ; Hinrich Schütze. 1. ed. Cham, 2024. pp. 137-150 (Springer Proceedings in Advanced Robotics (SPAR); 29).
Download
@inproceedings{f36b8ea6336940e7a525e36e6c475f3b,
title = "Quantifying Uncertainties of Contact Classifications in a Human-Robot Collaboration with Parallel Robots",
abstract = "In human-robot collaboration, unintentional physical contacts occur in the form of collisions and clamping, which must be detected and classified separately for a reaction. If certain collision or clamping situations are misclassified, reactions might occur that make the true contact case more dangerous. This work analyzes data-driven modeling based on physically modeled features like estimated external forces for clamping and collision classification with a real parallel robot. The prediction reliability of a feedforward neural network is investigated. Quantification of the classification uncertainty enables the distinction between safe versus unreliable classifications and optimal reactions like a retraction movement for collisions, structure opening for the clamping joint, and a fallback reaction in the form of a zero-g mode. This hypothesis is tested with experimental data of clamping and collision cases by analyzing dangerous misclassifications and then reducing them by the proposed uncertainty quantification. Finally, it is investigated how the approach of this work influences correctly classified clamping and collision scenarios.",
keywords = "cs.RO, cs.SY, eess.SY, parallel robots, human-robot collaboration, data-driven modeling",
author = "Aran Mohammad and Hendrik Muscheid and Moritz Schappler and Thomas Seel",
note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.; 16th International Workshop on Human-Friendly Robotics (HFR 2023), HFR 2023 ; Conference date: 20-09-2023 Through 21-09-2023",
year = "2024",
month = mar,
day = "10",
doi = "10.1007/978-3-031-55000-3_10",
language = "English",
isbn = "978-3-031-54999-1",
series = "Springer Proceedings in Advanced Robotics (SPAR)",
number = "29",
pages = "137--150",
editor = "Christina Piazza and Patricia Capsi-Morales and Luis Figueredo and Manuel Keppler and Hinrich Sch{\"u}tze",
booktitle = "Human-Friendly Robotics 2023",
edition = "1.",

}

Download

TY - GEN

T1 - Quantifying Uncertainties of Contact Classifications in a Human-Robot Collaboration with Parallel Robots

AU - Mohammad, Aran

AU - Muscheid, Hendrik

AU - Schappler, Moritz

AU - Seel, Thomas

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

PY - 2024/3/10

Y1 - 2024/3/10

N2 - In human-robot collaboration, unintentional physical contacts occur in the form of collisions and clamping, which must be detected and classified separately for a reaction. If certain collision or clamping situations are misclassified, reactions might occur that make the true contact case more dangerous. This work analyzes data-driven modeling based on physically modeled features like estimated external forces for clamping and collision classification with a real parallel robot. The prediction reliability of a feedforward neural network is investigated. Quantification of the classification uncertainty enables the distinction between safe versus unreliable classifications and optimal reactions like a retraction movement for collisions, structure opening for the clamping joint, and a fallback reaction in the form of a zero-g mode. This hypothesis is tested with experimental data of clamping and collision cases by analyzing dangerous misclassifications and then reducing them by the proposed uncertainty quantification. Finally, it is investigated how the approach of this work influences correctly classified clamping and collision scenarios.

AB - In human-robot collaboration, unintentional physical contacts occur in the form of collisions and clamping, which must be detected and classified separately for a reaction. If certain collision or clamping situations are misclassified, reactions might occur that make the true contact case more dangerous. This work analyzes data-driven modeling based on physically modeled features like estimated external forces for clamping and collision classification with a real parallel robot. The prediction reliability of a feedforward neural network is investigated. Quantification of the classification uncertainty enables the distinction between safe versus unreliable classifications and optimal reactions like a retraction movement for collisions, structure opening for the clamping joint, and a fallback reaction in the form of a zero-g mode. This hypothesis is tested with experimental data of clamping and collision cases by analyzing dangerous misclassifications and then reducing them by the proposed uncertainty quantification. Finally, it is investigated how the approach of this work influences correctly classified clamping and collision scenarios.

KW - cs.RO

KW - cs.SY

KW - eess.SY

KW - parallel robots

KW - human-robot collaboration

KW - data-driven modeling

UR - http://www.scopus.com/inward/record.url?scp=85188745666&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-55000-3_10

DO - 10.1007/978-3-031-55000-3_10

M3 - Conference contribution

SN - 978-3-031-54999-1

T3 - Springer Proceedings in Advanced Robotics (SPAR)

SP - 137

EP - 150

BT - Human-Friendly Robotics 2023

A2 - Piazza, Christina

A2 - Capsi-Morales, Patricia

A2 - Figueredo, Luis

A2 - Keppler, Manuel

A2 - Schütze, Hinrich

CY - Cham

T2 - 16th International Workshop on Human-Friendly Robotics (HFR 2023)

Y2 - 20 September 2023 through 21 September 2023

ER -

By the same author(s)