Learning interaction for collaborative tasks with probabilistic movement primitives

Guilherme Maeda, Marco Ewerton, Rudolf Lioutikov, Hani Ben Amor, Jan Peters, Gerhard Neumann

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

25 Citations (Scopus)

Abstract

This paper proposes a probabilistic framework based on movement primitives for robots that work in collaboration with a human coworker. Since the human coworker can execute a variety of unforeseen tasks a requirement of our system is that the robot assistant must be able to adapt and learn new skills on-demand, without the need of an expert programmer. Thus, this paper leverages on the framework of imitation learning and its application to human-robot interaction using the concept of Interaction Primitives (IPs). We introduce the use of Probabilistic Movement Primitives (ProMPs) to devise an interaction method that both recognizes the action of a human and generates the appropriate movement primitive of the robot assistant. We evaluate our method on experiments using a lightweight arm interacting with a human partner and also using motion capture trajectories of two humans assembling a box. The advantages of ProMPs in relation to the original formulation for interaction are exposed and compared.

Original languageEnglish (US)
Title of host publicationIEEE-RAS International Conference on Humanoid Robots
PublisherIEEE Computer Society
Pages527-534
Number of pages8
Volume2015-February
ISBN (Print)9781479971749
DOIs
StatePublished - Feb 12 2015
Externally publishedYes
Event2014 14th IEEE-RAS International Conference on Humanoid Robots, Humanoids 2014 - Madrid, Spain
Duration: Nov 18 2014Nov 20 2014

Other

Other2014 14th IEEE-RAS International Conference on Humanoid Robots, Humanoids 2014
CountrySpain
CityMadrid
Period11/18/1411/20/14

Fingerprint

Robots
Human robot interaction
Trajectories
Experiments

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Hardware and Architecture
  • Human-Computer Interaction
  • Electrical and Electronic Engineering

Cite this

Maeda, G., Ewerton, M., Lioutikov, R., Ben Amor, H., Peters, J., & Neumann, G. (2015). Learning interaction for collaborative tasks with probabilistic movement primitives. In IEEE-RAS International Conference on Humanoid Robots (Vol. 2015-February, pp. 527-534). [7041413] IEEE Computer Society. https://doi.org/10.1109/HUMANOIDS.2014.7041413

Learning interaction for collaborative tasks with probabilistic movement primitives. / Maeda, Guilherme; Ewerton, Marco; Lioutikov, Rudolf; Ben Amor, Hani; Peters, Jan; Neumann, Gerhard.

IEEE-RAS International Conference on Humanoid Robots. Vol. 2015-February IEEE Computer Society, 2015. p. 527-534 7041413.

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

Maeda, G, Ewerton, M, Lioutikov, R, Ben Amor, H, Peters, J & Neumann, G 2015, Learning interaction for collaborative tasks with probabilistic movement primitives. in IEEE-RAS International Conference on Humanoid Robots. vol. 2015-February, 7041413, IEEE Computer Society, pp. 527-534, 2014 14th IEEE-RAS International Conference on Humanoid Robots, Humanoids 2014, Madrid, Spain, 11/18/14. https://doi.org/10.1109/HUMANOIDS.2014.7041413
Maeda G, Ewerton M, Lioutikov R, Ben Amor H, Peters J, Neumann G. Learning interaction for collaborative tasks with probabilistic movement primitives. In IEEE-RAS International Conference on Humanoid Robots. Vol. 2015-February. IEEE Computer Society. 2015. p. 527-534. 7041413 https://doi.org/10.1109/HUMANOIDS.2014.7041413
Maeda, Guilherme ; Ewerton, Marco ; Lioutikov, Rudolf ; Ben Amor, Hani ; Peters, Jan ; Neumann, Gerhard. / Learning interaction for collaborative tasks with probabilistic movement primitives. IEEE-RAS International Conference on Humanoid Robots. Vol. 2015-February IEEE Computer Society, 2015. pp. 527-534
@inproceedings{66da6a505393418f8011858dd52d30cc,
title = "Learning interaction for collaborative tasks with probabilistic movement primitives",
abstract = "This paper proposes a probabilistic framework based on movement primitives for robots that work in collaboration with a human coworker. Since the human coworker can execute a variety of unforeseen tasks a requirement of our system is that the robot assistant must be able to adapt and learn new skills on-demand, without the need of an expert programmer. Thus, this paper leverages on the framework of imitation learning and its application to human-robot interaction using the concept of Interaction Primitives (IPs). We introduce the use of Probabilistic Movement Primitives (ProMPs) to devise an interaction method that both recognizes the action of a human and generates the appropriate movement primitive of the robot assistant. We evaluate our method on experiments using a lightweight arm interacting with a human partner and also using motion capture trajectories of two humans assembling a box. The advantages of ProMPs in relation to the original formulation for interaction are exposed and compared.",
author = "Guilherme Maeda and Marco Ewerton and Rudolf Lioutikov and {Ben Amor}, Hani and Jan Peters and Gerhard Neumann",
year = "2015",
month = "2",
day = "12",
doi = "10.1109/HUMANOIDS.2014.7041413",
language = "English (US)",
isbn = "9781479971749",
volume = "2015-February",
pages = "527--534",
booktitle = "IEEE-RAS International Conference on Humanoid Robots",
publisher = "IEEE Computer Society",

}

TY - GEN

T1 - Learning interaction for collaborative tasks with probabilistic movement primitives

AU - Maeda, Guilherme

AU - Ewerton, Marco

AU - Lioutikov, Rudolf

AU - Ben Amor, Hani

AU - Peters, Jan

AU - Neumann, Gerhard

PY - 2015/2/12

Y1 - 2015/2/12

N2 - This paper proposes a probabilistic framework based on movement primitives for robots that work in collaboration with a human coworker. Since the human coworker can execute a variety of unforeseen tasks a requirement of our system is that the robot assistant must be able to adapt and learn new skills on-demand, without the need of an expert programmer. Thus, this paper leverages on the framework of imitation learning and its application to human-robot interaction using the concept of Interaction Primitives (IPs). We introduce the use of Probabilistic Movement Primitives (ProMPs) to devise an interaction method that both recognizes the action of a human and generates the appropriate movement primitive of the robot assistant. We evaluate our method on experiments using a lightweight arm interacting with a human partner and also using motion capture trajectories of two humans assembling a box. The advantages of ProMPs in relation to the original formulation for interaction are exposed and compared.

AB - This paper proposes a probabilistic framework based on movement primitives for robots that work in collaboration with a human coworker. Since the human coworker can execute a variety of unforeseen tasks a requirement of our system is that the robot assistant must be able to adapt and learn new skills on-demand, without the need of an expert programmer. Thus, this paper leverages on the framework of imitation learning and its application to human-robot interaction using the concept of Interaction Primitives (IPs). We introduce the use of Probabilistic Movement Primitives (ProMPs) to devise an interaction method that both recognizes the action of a human and generates the appropriate movement primitive of the robot assistant. We evaluate our method on experiments using a lightweight arm interacting with a human partner and also using motion capture trajectories of two humans assembling a box. The advantages of ProMPs in relation to the original formulation for interaction are exposed and compared.

UR - http://www.scopus.com/inward/record.url?scp=84945190639&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84945190639&partnerID=8YFLogxK

U2 - 10.1109/HUMANOIDS.2014.7041413

DO - 10.1109/HUMANOIDS.2014.7041413

M3 - Conference contribution

SN - 9781479971749

VL - 2015-February

SP - 527

EP - 534

BT - IEEE-RAS International Conference on Humanoid Robots

PB - IEEE Computer Society

ER -