HyperNEAT for Locomotion Control in Modular Robots

E.W. Haasdijk, A.A. Rusu, A.E. Eiben

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

In an application where autonomous robots can amalgamate spontaneously
into arbitrary organisms, the individual robots cannot know a priori at
which location in an organism they will end up. If the organism is to be controlled
autonomously by the constituent robots, an evolutionary algorithm that
evolves the controllers can only develop a single genome that will have to suffice
for every individual robot. However, the robots should show different behaviour
depending on their position in an organism, meaning their phenotype should be
different depending on their location. In this paper, we demonstrate a solution
for this problem using the HyperNEAT generative encoding technique with differentiated
genome expression. We develop controllers for organism locomotion
with obstacle avoidance as a proof of concept. Finally, we identify promising
directions for further research.
Original languageEnglish
Pages (from-to)169-180
JournalLecture Notes in Computer Science
Volume6247
DOIs
Publication statusPublished - 2010
EventICES 2010 - Heidelberg
Duration: 6 Sep 20108 Sep 2010

Fingerprint

Modular robots
Locomotion
Robot
Robots
Controller
Autonomous Robots
Obstacle Avoidance
Controllers
Phenotype
Evolutionary Algorithms
Collision avoidance
Genome
Encoding
Evolutionary algorithms
Genes
Arbitrary
Demonstrate

Bibliographical note

Proceedings title: Evolvable Systems: From Biology to Hardware, proceedings of the 9th International Conference on Evolvable Systems (ICES 2010)
Publisher: Springer Verlag
Place of publication: Heidelberg
Editors: G Tempesti, A.M Tyrell, J.F Miller

Cite this

Haasdijk, E.W. ; Rusu, A.A. ; Eiben, A.E. / HyperNEAT for Locomotion Control in Modular Robots. In: Lecture Notes in Computer Science. 2010 ; Vol. 6247. pp. 169-180.
@article{2d02ec70829e48149c38fa4446dd3f1f,
title = "HyperNEAT for Locomotion Control in Modular Robots",
abstract = "In an application where autonomous robots can amalgamate spontaneouslyinto arbitrary organisms, the individual robots cannot know a priori atwhich location in an organism they will end up. If the organism is to be controlledautonomously by the constituent robots, an evolutionary algorithm thatevolves the controllers can only develop a single genome that will have to sufficefor every individual robot. However, the robots should show different behaviourdepending on their position in an organism, meaning their phenotype should bedifferent depending on their location. In this paper, we demonstrate a solutionfor this problem using the HyperNEAT generative encoding technique with differentiatedgenome expression. We develop controllers for organism locomotionwith obstacle avoidance as a proof of concept. Finally, we identify promisingdirections for further research.",
author = "E.W. Haasdijk and A.A. Rusu and A.E. Eiben",
note = "Proceedings title: Evolvable Systems: From Biology to Hardware, proceedings of the 9th International Conference on Evolvable Systems (ICES 2010) Publisher: Springer Verlag Place of publication: Heidelberg Editors: G Tempesti, A.M Tyrell, J.F Miller",
year = "2010",
doi = "10.1007/978-3-642-15323-5_15",
language = "English",
volume = "6247",
pages = "169--180",
journal = "Lecture Notes in Computer Science",
issn = "0302-9743",
publisher = "Springer Verlag",

}

HyperNEAT for Locomotion Control in Modular Robots. / Haasdijk, E.W.; Rusu, A.A.; Eiben, A.E.

In: Lecture Notes in Computer Science, Vol. 6247, 2010, p. 169-180.

Research output: Contribution to JournalArticleAcademicpeer-review

TY - JOUR

T1 - HyperNEAT for Locomotion Control in Modular Robots

AU - Haasdijk, E.W.

AU - Rusu, A.A.

AU - Eiben, A.E.

N1 - Proceedings title: Evolvable Systems: From Biology to Hardware, proceedings of the 9th International Conference on Evolvable Systems (ICES 2010) Publisher: Springer Verlag Place of publication: Heidelberg Editors: G Tempesti, A.M Tyrell, J.F Miller

PY - 2010

Y1 - 2010

N2 - In an application where autonomous robots can amalgamate spontaneouslyinto arbitrary organisms, the individual robots cannot know a priori atwhich location in an organism they will end up. If the organism is to be controlledautonomously by the constituent robots, an evolutionary algorithm thatevolves the controllers can only develop a single genome that will have to sufficefor every individual robot. However, the robots should show different behaviourdepending on their position in an organism, meaning their phenotype should bedifferent depending on their location. In this paper, we demonstrate a solutionfor this problem using the HyperNEAT generative encoding technique with differentiatedgenome expression. We develop controllers for organism locomotionwith obstacle avoidance as a proof of concept. Finally, we identify promisingdirections for further research.

AB - In an application where autonomous robots can amalgamate spontaneouslyinto arbitrary organisms, the individual robots cannot know a priori atwhich location in an organism they will end up. If the organism is to be controlledautonomously by the constituent robots, an evolutionary algorithm thatevolves the controllers can only develop a single genome that will have to sufficefor every individual robot. However, the robots should show different behaviourdepending on their position in an organism, meaning their phenotype should bedifferent depending on their location. In this paper, we demonstrate a solutionfor this problem using the HyperNEAT generative encoding technique with differentiatedgenome expression. We develop controllers for organism locomotionwith obstacle avoidance as a proof of concept. Finally, we identify promisingdirections for further research.

U2 - 10.1007/978-3-642-15323-5_15

DO - 10.1007/978-3-642-15323-5_15

M3 - Article

VL - 6247

SP - 169

EP - 180

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -