Learning Affordances for Assistive Robots

Mohan Sridharan; Ben Meadows

doi:10.1007/978-3-319-70022-9_1

Learning Affordances for Assistive Robots

Mohan Sridharan, Ben Meadows

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

186 Downloads (Pure)

Abstract

This paper describes an architecture that enables a robot to represent,
reason about, and learn affordances. Specifically, Answer Set Prolog is used
to represent and reason with incomplete domain knowledge that includes affordances modeled as relations between attributes of the robot and the object(s) in the context of specific actions. The learning of affordance relations from observations obtained through reactive execution or active exploration is formulated as a reinforcement learning problem. A sampling-based approach and decision-tree regression with the underlying relational representation are used to obtain generic affordance relations that are added to the Answer Set Prolog program for subsequent reasoning. The capabilities of this architecture are illustrated and evaluated in the context of a simulated robot assisting humans in an indoor domain.

Original language	English
Title of host publication	Social Robotics
Subtitle of host publication	9th International Conference, ICSR 2017, Tsukuba, Japan, November 22-24, 2017, Proceedings
Editors	A. Kheddar, E. Yoshida, S.S. Ge, K. Suzuki, J.-J. Cabibihan, F. Eyssel, H. He
Publisher	Springer
Chapter	1
Pages	1-11
ISBN (Electronic)	978-3-319-70022-9
ISBN (Print)	978-3-319-70021-2
DOIs	https://doi.org/10.1007/978-3-319-70022-9_1
Publication status	Published - 24 Oct 2017
Event	9th International Conference on Social Robotics 2017 (ICSR) - Tsukuba, Japan Duration: 22 Nov 2017 → 24 Nov 2017

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
Volume	10652
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	9th International Conference on Social Robotics 2017 (ICSR)
Country/Territory	Japan
City	Tsukuba
Period	22/11/17 → 24/11/17

Access to Document

10.1007/978-3-319-70022-9_1

Sridharan_Learning_Affordances_LNCS
This is a post-peer-review, pre-copyedit version of an article published in Lecture Notes in Computer Science. The final authenticated version is available online at: http://dx.doi.org/10.1007/978-3-319-70022-9_1
Accepted author manuscript, 476 KBLicence: None: All rights reserved

Cite this

Sridharan, M., & Meadows, B. (2017). Learning Affordances for Assistive Robots. In A. Kheddar, E. Yoshida, S. S. Ge, K. Suzuki, J-J. Cabibihan, F. Eyssel, & H. He (Eds.), Social Robotics: 9th International Conference, ICSR 2017, Tsukuba, Japan, November 22-24, 2017, Proceedings (pp. 1-11). (Lecture Notes in Computer Science; Vol. 10652). Springer. https://doi.org/10.1007/978-3-319-70022-9_1

@inproceedings{99819f8c21b945c08c0ef278bb395ab1,

title = "Learning Affordances for Assistive Robots",

abstract = "This paper describes an architecture that enables a robot to represent,reason about, and learn affordances. Specifically, Answer Set Prolog is usedto represent and reason with incomplete domain knowledge that includes affordances modeled as relations between attributes of the robot and the object(s) in the context of specific actions. The learning of affordance relations from observations obtained through reactive execution or active exploration is formulated as a reinforcement learning problem. A sampling-based approach and decision-tree regression with the underlying relational representation are used to obtain generic affordance relations that are added to the Answer Set Prolog program for subsequent reasoning. The capabilities of this architecture are illustrated and evaluated in the context of a simulated robot assisting humans in an indoor domain.",

author = "Mohan Sridharan and Ben Meadows",

year = "2017",

month = oct,

day = "24",

doi = "10.1007/978-3-319-70022-9_1",

language = "English",

isbn = "978-3-319-70021-2",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "1--11",

editor = "A. Kheddar and E. Yoshida and S.S. Ge and K. Suzuki and J.-J. Cabibihan and F. Eyssel and H. He",

booktitle = "Social Robotics",

note = "9th International Conference on Social Robotics 2017 (ICSR) ; Conference date: 22-11-2017 Through 24-11-2017",

}

Sridharan, M & Meadows, B 2017, Learning Affordances for Assistive Robots. in A Kheddar, E Yoshida, SS Ge, K Suzuki, J-J Cabibihan, F Eyssel & H He (eds), Social Robotics: 9th International Conference, ICSR 2017, Tsukuba, Japan, November 22-24, 2017, Proceedings. Lecture Notes in Computer Science, vol. 10652, Springer, pp. 1-11, 9th International Conference on Social Robotics 2017 (ICSR), Tsukuba, Japan, 22/11/17. https://doi.org/10.1007/978-3-319-70022-9_1

Learning Affordances for Assistive Robots. / Sridharan, Mohan; Meadows, Ben.
Social Robotics: 9th International Conference, ICSR 2017, Tsukuba, Japan, November 22-24, 2017, Proceedings. ed. / A. Kheddar; E. Yoshida; S.S. Ge; K. Suzuki; J.-J. Cabibihan; F. Eyssel; H. He. Springer, 2017. p. 1-11 (Lecture Notes in Computer Science; Vol. 10652).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Learning Affordances for Assistive Robots

AU - Sridharan, Mohan

AU - Meadows, Ben

PY - 2017/10/24

Y1 - 2017/10/24

N2 - This paper describes an architecture that enables a robot to represent,reason about, and learn affordances. Specifically, Answer Set Prolog is usedto represent and reason with incomplete domain knowledge that includes affordances modeled as relations between attributes of the robot and the object(s) in the context of specific actions. The learning of affordance relations from observations obtained through reactive execution or active exploration is formulated as a reinforcement learning problem. A sampling-based approach and decision-tree regression with the underlying relational representation are used to obtain generic affordance relations that are added to the Answer Set Prolog program for subsequent reasoning. The capabilities of this architecture are illustrated and evaluated in the context of a simulated robot assisting humans in an indoor domain.

AB - This paper describes an architecture that enables a robot to represent,reason about, and learn affordances. Specifically, Answer Set Prolog is usedto represent and reason with incomplete domain knowledge that includes affordances modeled as relations between attributes of the robot and the object(s) in the context of specific actions. The learning of affordance relations from observations obtained through reactive execution or active exploration is formulated as a reinforcement learning problem. A sampling-based approach and decision-tree regression with the underlying relational representation are used to obtain generic affordance relations that are added to the Answer Set Prolog program for subsequent reasoning. The capabilities of this architecture are illustrated and evaluated in the context of a simulated robot assisting humans in an indoor domain.

U2 - 10.1007/978-3-319-70022-9_1

DO - 10.1007/978-3-319-70022-9_1

M3 - Conference contribution

SN - 978-3-319-70021-2

T3 - Lecture Notes in Computer Science

SP - 1

EP - 11

BT - Social Robotics

A2 - Kheddar, A.

A2 - Yoshida, E.

A2 - Ge, S.S.

A2 - Suzuki, K.

A2 - Cabibihan, J.-J.

A2 - Eyssel, F.

A2 - He, H.

PB - Springer

T2 - 9th International Conference on Social Robotics 2017 (ICSR)

Y2 - 22 November 2017 through 24 November 2017

ER -

Learning Affordances for Assistive Robots

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this