Data · collection · 2017
Supplementary material from "Evolutionary online behaviour learning and adaptation in real robots"
Listed in DataCite
Online evolution of behavioural control on real robots is an open-ended approach to autonomous learning and adaptation: robots have the potential to automatically learn new tasks and to adapt to changes in environmental conditions, or to failures in sensors and/or actuators.
Description
However, studies have so far almost exclusively been carried out in simulation because evolution in real hardware has required several days or weeks to produce capable robots.
In this article, we successfully evolve neural network-based controllers in real robotic hardware to solve two single-robot tasks and one collective robotics task. Controllers are evolved either from random solutions or from solutions pre-evolved in simulation. In all cases, capable solutions are found in a timely manner (1 h or less).
Read the rest (2 more)
Results show that more accurate simulations may lead to higher-performing controllers, and that completing the optimization process in real robots is meaningful, even if solutions found in simulation differ from solutions in reality. We, furthermore, demonstrate for the first time the adaptive capabilities of online evolution in real robotic hardware, including robots able to overcome faults injected in the motors of multiple units simultaneously, and to modify their behaviour in response to changes in the task requirements.
We conclude by assessing the contribution of each algorithmic component on the performance of the underlying evolutionary algorithm.
Links
Where it is published
- Repository landing page figshare.com/collections/Supplementary_material_from_Evolutionary_online_be… ↗
landing page · from DataCite
- DOI doi.org/10.6084/m9.figshare.c.3827944 ↗
DOI / persistent id · from DataCite
Documentation and papers
- CC BY 4.0 creativecommons.org/licenses/by/4.0 ↗
license · from DataCite
- IsSupplementTo 10.1098/rsos.160938 doi.org/10.1098/rsos.160938 ↗
publication · from DataCite
Catalogue records · 2
- DataCite API api.datacite.org/dois/10.6084/m9.figshare.c.3827944 ↗
metadata API · from DataCite
- DataCite Commons commons.datacite.org/doi.org/10.6084/m9.figshare.c.3827944 ↗
catalogue entry · from DataCite
Topics
- Stated by source
- Computer and information sciences · Electrical engineering, electronic engineering, information engineering
- Inferred from text
- Simulation 75%
Related
Provenance · 1 source records, 10 field assertions
| Source | Key | Last seen | Raw |
|---|---|---|---|
| DataCite | 10.6084/m9.figshare.c.3827944 | 14 d ago | JSON v1 |
| Field | Assertion | Extractor | Evidence |
|---|---|---|---|
| access_level | source · DataCite | connector:datacite@1.0.0 | /data/attributes/rightsList |
| concepts[field].fos:computer-and-information-sciences | source · DataCite | connector:datacite@1.0.0 | |
| concepts[field].fos:electrical-engineering-electronic-engineering-information-engineering | source · DataCite | connector:datacite@1.0.0 | |
| concepts[method].local:method:simulation | enrichment · DataCite | keyword-concept-rules@1.0.0 | title+description (75%) |
| created_date | source · DataCite | connector:datacite@1.0.0 | |
| description | source · DataCite | connector:datacite@1.0.0 | /data/attributes/descriptions |
| license | source · DataCite | connector:datacite@1.0.0 | /data/attributes/rightsList |
| publication_date | source · DataCite | connector:datacite@1.0.0 | /data/attributes/dates |
| title | source · DataCite | connector:datacite@1.0.0 | /data/attributes/titles/0/title |
| updated_date | source · DataCite | connector:datacite@1.0.0 |