From real-time adaptation to social learning in robot ecosystems
While evolutionary robotics can create novel morphologies and controllers that are well-adapted to their environments, learning is still the most efficient way to adapt to changes that occur on shorter time scales. Learning proposals for evolving robots to date have focused on new individuals either...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Frontiers Media S.A.
2023-10-01
|
Series: | Frontiers in Robotics and AI |
Subjects: | |
Online Access: | https://www.frontiersin.org/articles/10.3389/frobt.2023.1232708/full |
_version_ | 1797666335193300992 |
---|---|
author | Alex Szorkovszky Alex Szorkovszky Frank Veenstra Frank Veenstra Kyrre Glette Kyrre Glette |
author_facet | Alex Szorkovszky Alex Szorkovszky Frank Veenstra Frank Veenstra Kyrre Glette Kyrre Glette |
author_sort | Alex Szorkovszky |
collection | DOAJ |
description | While evolutionary robotics can create novel morphologies and controllers that are well-adapted to their environments, learning is still the most efficient way to adapt to changes that occur on shorter time scales. Learning proposals for evolving robots to date have focused on new individuals either learning a controller from scratch, or building on the experience of direct ancestors and/or robots with similar configurations. Here we propose and demonstrate a novel means for social learning of gait patterns, based on sensorimotor synchronization. Using movement patterns of other robots as input can drive nonlinear decentralized controllers such as CPGs into new limit cycles, hence encouraging diversity of movement patterns. Stable autonomous controllers can then be locked in, which we demonstrate using a quasi-Hebbian feedback scheme. We propose that in an ecosystem of robots evolving in a heterogeneous environment, such a scheme may allow for the emergence of generalist task-solvers from a population of specialists. |
first_indexed | 2024-03-11T19:58:07Z |
format | Article |
id | doaj.art-0b473cdc01394eca89fbe9490058ff87 |
institution | Directory Open Access Journal |
issn | 2296-9144 |
language | English |
last_indexed | 2024-03-11T19:58:07Z |
publishDate | 2023-10-01 |
publisher | Frontiers Media S.A. |
record_format | Article |
series | Frontiers in Robotics and AI |
spelling | doaj.art-0b473cdc01394eca89fbe9490058ff872023-10-04T14:03:04ZengFrontiers Media S.A.Frontiers in Robotics and AI2296-91442023-10-011010.3389/frobt.2023.12327081232708From real-time adaptation to social learning in robot ecosystemsAlex Szorkovszky0Alex Szorkovszky1Frank Veenstra2Frank Veenstra3Kyrre Glette4Kyrre Glette5RITMO Centre for Interdisciplinary Studies in Rhythm, Time and Motion, University of Oslo, Oslo, NorwayDepartment of Informatics, University of Oslo, Oslo, NorwayRITMO Centre for Interdisciplinary Studies in Rhythm, Time and Motion, University of Oslo, Oslo, NorwayDepartment of Informatics, University of Oslo, Oslo, NorwayRITMO Centre for Interdisciplinary Studies in Rhythm, Time and Motion, University of Oslo, Oslo, NorwayDepartment of Informatics, University of Oslo, Oslo, NorwayWhile evolutionary robotics can create novel morphologies and controllers that are well-adapted to their environments, learning is still the most efficient way to adapt to changes that occur on shorter time scales. Learning proposals for evolving robots to date have focused on new individuals either learning a controller from scratch, or building on the experience of direct ancestors and/or robots with similar configurations. Here we propose and demonstrate a novel means for social learning of gait patterns, based on sensorimotor synchronization. Using movement patterns of other robots as input can drive nonlinear decentralized controllers such as CPGs into new limit cycles, hence encouraging diversity of movement patterns. Stable autonomous controllers can then be locked in, which we demonstrate using a quasi-Hebbian feedback scheme. We propose that in an ecosystem of robots evolving in a heterogeneous environment, such a scheme may allow for the emergence of generalist task-solvers from a population of specialists.https://www.frontiersin.org/articles/10.3389/frobt.2023.1232708/fullsocial learningevolutionary roboticsentrainmentcentral pattern generatorcultural evolution |
spellingShingle | Alex Szorkovszky Alex Szorkovszky Frank Veenstra Frank Veenstra Kyrre Glette Kyrre Glette From real-time adaptation to social learning in robot ecosystems Frontiers in Robotics and AI social learning evolutionary robotics entrainment central pattern generator cultural evolution |
title | From real-time adaptation to social learning in robot ecosystems |
title_full | From real-time adaptation to social learning in robot ecosystems |
title_fullStr | From real-time adaptation to social learning in robot ecosystems |
title_full_unstemmed | From real-time adaptation to social learning in robot ecosystems |
title_short | From real-time adaptation to social learning in robot ecosystems |
title_sort | from real time adaptation to social learning in robot ecosystems |
topic | social learning evolutionary robotics entrainment central pattern generator cultural evolution |
url | https://www.frontiersin.org/articles/10.3389/frobt.2023.1232708/full |
work_keys_str_mv | AT alexszorkovszky fromrealtimeadaptationtosociallearninginrobotecosystems AT alexszorkovszky fromrealtimeadaptationtosociallearninginrobotecosystems AT frankveenstra fromrealtimeadaptationtosociallearninginrobotecosystems AT frankveenstra fromrealtimeadaptationtosociallearninginrobotecosystems AT kyrreglette fromrealtimeadaptationtosociallearninginrobotecosystems AT kyrreglette fromrealtimeadaptationtosociallearninginrobotecosystems |