Measuring and Modelling Delays in Robot Manipulators for Temporally Precise Control using Machine Learning.

Thomas Timm Andersen, Heni Ben Amor, Nils Axel Andersen, Ole Ravn

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

652 Downloads (Pure)

Abstract

Latencies and delays play an important role in temporally precise robot control. During dynamic tasks in particular, a robot has to account for inherent delays to reach manipulated objects in time. The different types of occurring delays are typically convoluted and thereby hard to measure and separate. In this paper, we present a data-driven methodology for separating and modelling inherent delays during robot control. We show how both actuation and response delays can be modelled using modern machine learning methods. The resulting models can be used to predict the delays as well as the uncertainty of the prediction. Experiments on two widely used robot platforms show significant actuation and response delays in standard control loops. Predictive models can, therefore, be used to reason about expected delays and improve temporal accuracy during control. The approach can easily be used on different robot platforms.
Original languageEnglish
Title of host publicationProceedings of IEEE ICMLA'15
Number of pages8
PublisherIEEE
Publication date2015
Publication statusPublished - 2015
Event14th International Conference on Machine Learning and Applications - Miami, Florida, United States
Duration: 9 Dec 201511 Dec 2015

Conference

Conference14th International Conference on Machine Learning and Applications
CountryUnited States
CityMiami, Florida
Period09/12/201511/12/2015

Keywords

  • Robot control
  • Automation
  • Machine learning algorithms

Cite this

@inproceedings{539bcee9fce44cb2b2617f5db5217590,
title = "Measuring and Modelling Delays in Robot Manipulators for Temporally Precise Control using Machine Learning.",
abstract = "Latencies and delays play an important role in temporally precise robot control. During dynamic tasks in particular, a robot has to account for inherent delays to reach manipulated objects in time. The different types of occurring delays are typically convoluted and thereby hard to measure and separate. In this paper, we present a data-driven methodology for separating and modelling inherent delays during robot control. We show how both actuation and response delays can be modelled using modern machine learning methods. The resulting models can be used to predict the delays as well as the uncertainty of the prediction. Experiments on two widely used robot platforms show significant actuation and response delays in standard control loops. Predictive models can, therefore, be used to reason about expected delays and improve temporal accuracy during control. The approach can easily be used on different robot platforms.",
keywords = "Robot control, Automation, Machine learning algorithms",
author = "Andersen, {Thomas Timm} and Amor, {Heni Ben} and Andersen, {Nils Axel} and Ole Ravn",
year = "2015",
language = "English",
booktitle = "Proceedings of IEEE ICMLA'15",
publisher = "IEEE",
address = "United States",

}

Andersen, TT, Amor, HB, Andersen, NA & Ravn, O 2015, Measuring and Modelling Delays in Robot Manipulators for Temporally Precise Control using Machine Learning. in Proceedings of IEEE ICMLA'15. IEEE, 14th International Conference on Machine Learning and Applications , Miami, Florida, United States, 09/12/2015.

Measuring and Modelling Delays in Robot Manipulators for Temporally Precise Control using Machine Learning. / Andersen, Thomas Timm; Amor, Heni Ben; Andersen, Nils Axel; Ravn, Ole.

Proceedings of IEEE ICMLA'15. IEEE, 2015.

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

TY - GEN

T1 - Measuring and Modelling Delays in Robot Manipulators for Temporally Precise Control using Machine Learning.

AU - Andersen, Thomas Timm

AU - Amor, Heni Ben

AU - Andersen, Nils Axel

AU - Ravn, Ole

PY - 2015

Y1 - 2015

N2 - Latencies and delays play an important role in temporally precise robot control. During dynamic tasks in particular, a robot has to account for inherent delays to reach manipulated objects in time. The different types of occurring delays are typically convoluted and thereby hard to measure and separate. In this paper, we present a data-driven methodology for separating and modelling inherent delays during robot control. We show how both actuation and response delays can be modelled using modern machine learning methods. The resulting models can be used to predict the delays as well as the uncertainty of the prediction. Experiments on two widely used robot platforms show significant actuation and response delays in standard control loops. Predictive models can, therefore, be used to reason about expected delays and improve temporal accuracy during control. The approach can easily be used on different robot platforms.

AB - Latencies and delays play an important role in temporally precise robot control. During dynamic tasks in particular, a robot has to account for inherent delays to reach manipulated objects in time. The different types of occurring delays are typically convoluted and thereby hard to measure and separate. In this paper, we present a data-driven methodology for separating and modelling inherent delays during robot control. We show how both actuation and response delays can be modelled using modern machine learning methods. The resulting models can be used to predict the delays as well as the uncertainty of the prediction. Experiments on two widely used robot platforms show significant actuation and response delays in standard control loops. Predictive models can, therefore, be used to reason about expected delays and improve temporal accuracy during control. The approach can easily be used on different robot platforms.

KW - Robot control

KW - Automation

KW - Machine learning algorithms

M3 - Article in proceedings

BT - Proceedings of IEEE ICMLA'15

PB - IEEE

ER -