English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

We generalize the well-studied problem of gait learning in modular robots intwo dimensions. Firstly, we address locomotion in a given target direction thatgoes beyond learning a typical undirected gait. Secondly, rather than studyingone fixed robot morphology we consider a test suite of different modularrobots. This study is based on our interest in evolutionary robot systems whereboth morphologies and controllers evolve. In such a system, newborn robots haveto learn to control their own body that is a random combination of the bodiesof the parents. We apply and compare two learning algorithms, Bayesianoptimization and HyperNEAT. The results of the experiments in simulation showthat both methods successfully learn good controllers, but Bayesianoptimization is more effective and efficient. We validate the best learnedcontrollers by constructing three robots from the test suite in the real worldand observe their fitness and actual trajectories. The obtained resultsindicate a reality gap that depends on the controllers and the shape of therobots, but overall the trajectories are adequate and follow the targetdirections successfully.

Learning Directed Locomotion in Modular Robots with Evolvable Morphologies