Local Models for Data Driven Inverse Kinematics of Soft Robots

Fredrik Dalland Holsten; Sune Darkner; Morten Pol  Engell-Nørregård; Kenny Erleben

doi:10.2312/sca.20181186

Local Models for Data Driven Inverse Kinematics of Soft Robots

Fredrik Dalland Holsten, Sune Darkner, Morten Pol Engell-Nørregård, Kenny Erleben

Department of Computer Science

Abstract

Soft robots are attractive because they have the potential of being safer, faster and cheaper than traditional rigid robots. If we can predict the shape of a soft robot for a given set of control parameters, then we can solve the inverse problem: to find an optimal set of control parameters for a given shape. This work takes a data-driven approach to create multiple local inverse models. This has two benefits: (1) We overcome the reality gap and (2) we gain performance and naive parallelism from using local models. Furthermore, we empirically prove that our approach outperforms a higher order global model. Eurographics Proceedingsc 2018 The Eurographics Association.

Original language	English
Title of host publication	Eurographics/ ACM SIGGRAPH Symposium on Computer Animation - Posters
Editors	Melina Skouras
Number of pages	1
Publisher	The Eurographics Association
Publication date	2018
ISBN (Print)	978-3-03868-070-3
DOIs	https://doi.org/10.2312/sca.20181186
Publication status	Published - 2018
Event	167h ACM SIGGRAPH / Eurographics Symposium on Computer Animation: SCA-18 - Paris, France Duration: 11 Jul 2018 → 13 Jul 2018

Conference

Conference	167h ACM SIGGRAPH / Eurographics Symposium on Computer Animation
Country/Territory	France
City	Paris
Period	11/07/2018 → 13/07/2018

Access to Document

10.2312/sca.20181186

Cite this

Local Models for Data Driven Inverse Kinematics of Soft Robots. / Holsten, Fredrik Dalland; Darkner, Sune; Engell-Nørregård, Morten Pol et al.

Eurographics/ ACM SIGGRAPH Symposium on Computer Animation - Posters. ed. / Melina Skouras. The Eurographics Association, 2018.

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Holsten, FD, Darkner, S, Engell-Nørregård, MP & Erleben, K 2018, Local Models for Data Driven Inverse Kinematics of Soft Robots. in M Skouras (ed.), Eurographics/ ACM SIGGRAPH Symposium on Computer Animation - Posters. The Eurographics Association, 167h ACM SIGGRAPH / Eurographics Symposium on Computer Animation, Paris, France, 11/07/2018. https://doi.org/10.2312/sca.20181186

@inproceedings{cb728af3b42c4ccfac8dbce1337ff93a,

title = "Local Models for Data Driven Inverse Kinematics of Soft Robots",

abstract = "Soft robots are attractive because they have the potential of being safer, faster and cheaper than traditional rigid robots. If we can predict the shape of a soft robot for a given set of control parameters, then we can solve the inverse problem: to find an optimal set of control parameters for a given shape. This work takes a data-driven approach to create multiple local inverse models. This has two benefits: (1) We overcome the reality gap and (2) we gain performance and naive parallelism from using local models. Furthermore, we empirically prove that our approach outperforms a higher order global model. Eurographics Proceedingsc 2018 The Eurographics Association.",

author = "Holsten, {Fredrik Dalland} and Sune Darkner and Engell-N{\o}rreg{\aa}rd, {Morten Pol} and Kenny Erleben",

year = "2018",

doi = "10.2312/sca.20181186",

language = "English",

isbn = "978-3-03868-070-3",

editor = "Melina Skouras",

booktitle = "Eurographics/ ACM SIGGRAPH Symposium on Computer Animation - Posters",

publisher = "The Eurographics Association",

note = "167h ACM SIGGRAPH / Eurographics Symposium on Computer Animation : SCA-18 ; Conference date: 11-07-2018 Through 13-07-2018",

}

TY - GEN

T1 - Local Models for Data Driven Inverse Kinematics of Soft Robots

AU - Holsten, Fredrik Dalland

AU - Darkner, Sune

AU - Engell-Nørregård, Morten Pol

AU - Erleben, Kenny

PY - 2018

Y1 - 2018

N2 - Soft robots are attractive because they have the potential of being safer, faster and cheaper than traditional rigid robots. If we can predict the shape of a soft robot for a given set of control parameters, then we can solve the inverse problem: to find an optimal set of control parameters for a given shape. This work takes a data-driven approach to create multiple local inverse models. This has two benefits: (1) We overcome the reality gap and (2) we gain performance and naive parallelism from using local models. Furthermore, we empirically prove that our approach outperforms a higher order global model. Eurographics Proceedingsc 2018 The Eurographics Association.

AB - Soft robots are attractive because they have the potential of being safer, faster and cheaper than traditional rigid robots. If we can predict the shape of a soft robot for a given set of control parameters, then we can solve the inverse problem: to find an optimal set of control parameters for a given shape. This work takes a data-driven approach to create multiple local inverse models. This has two benefits: (1) We overcome the reality gap and (2) we gain performance and naive parallelism from using local models. Furthermore, we empirically prove that our approach outperforms a higher order global model. Eurographics Proceedingsc 2018 The Eurographics Association.

U2 - 10.2312/sca.20181186

DO - 10.2312/sca.20181186

M3 - Article in proceedings

SN - 978-3-03868-070-3

BT - Eurographics/ ACM SIGGRAPH Symposium on Computer Animation - Posters

A2 - Skouras, Melina

PB - The Eurographics Association

T2 - 167h ACM SIGGRAPH / Eurographics Symposium on Computer Animation

Y2 - 11 July 2018 through 13 July 2018

ER -

Local Models for Data Driven Inverse Kinematics of Soft Robots

Abstract

Conference

Access to Document

Fingerprint

Cite this