A survey of robot learning from demonstration

Brenna D. Argall; Sonia Chernova; Manuela Veloso; Brett Browning

doi:10.1016/j.robot.2008.10.024

A survey of robot learning from demonstration

Brenna D. Argall^*, Sonia Chernova, Manuela Veloso, Brett Browning

^*Corresponding author for this work

Computer Science

Research output: Contribution to journal › Article › peer-review

2603 Scopus citations

Abstract

We present a comprehensive survey of robot Learning from Demonstration (LfD), a technique that develops policies from example state to action mappings. We introduce the LfD design choices in terms of demonstrator, problem space, policy derivation and performance, and contribute the foundations for a structure in which to categorize LfD research. Specifically, we analyze and categorize the multiple ways in which examples are gathered, ranging from teleoperation to imitation, as well as the various techniques for policy derivation, including matching functions, dynamics models and plans. To conclude we discuss LfD limitations and related promising areas for future research.

Original language	English (US)
Pages (from-to)	469-483
Number of pages	15
Journal	Robotics and Autonomous Systems
Volume	57
Issue number	5
DOIs	https://doi.org/10.1016/j.robot.2008.10.024
State	Published - May 31 2009

Funding

This research is partly sponsored by the Boeing Corporation under Grant No. CMU-BA-GTA-1, BBNT Solutions under subcontract No. 950008572, via prime Air Force contract No. SA-8650-06-C-7606, and the Qatar Foundation for Education, Science and Community Development. The views and conclusions contained in this document are solely those of the authors.

Keywords

Autonomous systems
Learning from demonstration
Machine learning
Robotics

ASJC Scopus subject areas

Control and Systems Engineering
Software
General Mathematics
Computer Science Applications

Access to Document

10.1016/j.robot.2008.10.024

Cite this

@article{12e2908dea004005a257707a1ad43237,

title = "A survey of robot learning from demonstration",

abstract = "We present a comprehensive survey of robot Learning from Demonstration (LfD), a technique that develops policies from example state to action mappings. We introduce the LfD design choices in terms of demonstrator, problem space, policy derivation and performance, and contribute the foundations for a structure in which to categorize LfD research. Specifically, we analyze and categorize the multiple ways in which examples are gathered, ranging from teleoperation to imitation, as well as the various techniques for policy derivation, including matching functions, dynamics models and plans. To conclude we discuss LfD limitations and related promising areas for future research.",

keywords = "Autonomous systems, Learning from demonstration, Machine learning, Robotics",

author = "Argall, {Brenna D.} and Sonia Chernova and Manuela Veloso and Brett Browning",

note = "Funding Information: This research is partly sponsored by the Boeing Corporation under Grant No. CMU-BA-GTA-1, BBNT Solutions under subcontract No. 950008572, via prime Air Force contract No. SA-8650-06-C-7606, and the Qatar Foundation for Education, Science and Community Development. The views and conclusions contained in this document are solely those of the authors. ",

year = "2009",

month = may,

day = "31",

doi = "10.1016/j.robot.2008.10.024",

language = "English (US)",

volume = "57",

pages = "469--483",

journal = "Robotics and Autonomous Systems",

issn = "0921-8890",

publisher = "Elsevier B.V.",

number = "5",

}

TY - JOUR

T1 - A survey of robot learning from demonstration

AU - Argall, Brenna D.

AU - Chernova, Sonia

AU - Veloso, Manuela

AU - Browning, Brett

N1 - Funding Information: This research is partly sponsored by the Boeing Corporation under Grant No. CMU-BA-GTA-1, BBNT Solutions under subcontract No. 950008572, via prime Air Force contract No. SA-8650-06-C-7606, and the Qatar Foundation for Education, Science and Community Development. The views and conclusions contained in this document are solely those of the authors.

PY - 2009/5/31

Y1 - 2009/5/31

N2 - We present a comprehensive survey of robot Learning from Demonstration (LfD), a technique that develops policies from example state to action mappings. We introduce the LfD design choices in terms of demonstrator, problem space, policy derivation and performance, and contribute the foundations for a structure in which to categorize LfD research. Specifically, we analyze and categorize the multiple ways in which examples are gathered, ranging from teleoperation to imitation, as well as the various techniques for policy derivation, including matching functions, dynamics models and plans. To conclude we discuss LfD limitations and related promising areas for future research.

AB - We present a comprehensive survey of robot Learning from Demonstration (LfD), a technique that develops policies from example state to action mappings. We introduce the LfD design choices in terms of demonstrator, problem space, policy derivation and performance, and contribute the foundations for a structure in which to categorize LfD research. Specifically, we analyze and categorize the multiple ways in which examples are gathered, ranging from teleoperation to imitation, as well as the various techniques for policy derivation, including matching functions, dynamics models and plans. To conclude we discuss LfD limitations and related promising areas for future research.

KW - Autonomous systems

KW - Learning from demonstration

KW - Machine learning

KW - Robotics

UR - http://www.scopus.com/inward/record.url?scp=63149159130&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=63149159130&partnerID=8YFLogxK

U2 - 10.1016/j.robot.2008.10.024

DO - 10.1016/j.robot.2008.10.024

M3 - Article

AN - SCOPUS:63149159130

SN - 0921-8890

VL - 57

SP - 469

EP - 483

JO - Robotics and Autonomous Systems

JF - Robotics and Autonomous Systems

IS - 5

ER -

A survey of robot learning from demonstration

Abstract

Funding

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this