Skip to content

This is the official repository for the GECCO (Genetic and Evolutionary Computation Conference, 2022) paper "A Comparative Analysis on Genome Pleiotropy for Evolved Soft Robots". This repository hosts all the code necessary to replicate the experiments. This work was carried out at Ghent University IDLab-AIRO.

License

Notifications You must be signed in to change notification settings

Co-Evolve/genome-pleiotropy

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

A Comparative Analysis on Genome Pleiotropy for Evolved Soft Robots

PDF Abstract: Biological evolution shapes the body and brain of living creatures together over time. By contrast, in evolutionary robotics, the co-optimization of these subsystems remains challenging. Conflicting mutations cause dissociation between morphology and control, which leads to premature convergence. Recent works have proposed algorithmic modifications to mitigate the impact of conflicting mutations. However, the importance of genetic design remains underexposed. Current approaches are divided between a single, pleiotropic genetic encoding and two isolated encodings representing morphology and control. This design choice is commonly made ad hoc, causing a lack of consistency for practitioners. To standardize this design, we performed a comparative analysis between these two configurations on a soft robot locomotion task. Additionally, we incorporated two currently unexplored alternatives that drive these configurations to their logical extremes. Our results demonstrate that pleiotropic representations yield superior performance in fitness and robustness towards premature convergence. Moreover, we showcase the importance of shared structure in the pleiotropic representation of robot morphology and control to achieve this performance gain. These findings provide valuable insights into genetic encoding design, which supply practitioners with a theoretical foundation to pursue efficient brain-body co-optimization.

genome_pleiotropy.mp4

Usage

Create the anaconda environment:

conda env create -f environment.yml

Run an experiment:

python -m pleiotropy.main --name {experiment_name} --params pleiotropy/experiment_configs/{configuration}

We refer to pleiotropy/experiment_configs/ for the hyperparameters used in every configuration. Metrics are logged and automatically saved on the cloud via Weights & Biases. Genomes are saved locally.

Citation

@inproceedings{10.1145/3520304.3528977,
author = {Marzougui, Dries and Biondina, Matthijs and wyffels, Francis},
title = {A Comparative Analysis on Genome Pleiotropy for Evolved Soft Robots},
year = {2022},
publisher = {Association for Computing Machinery},
url = {https://doi.org/10.1145/3520304.3528977},
doi = {10.1145/3520304.3528977},
booktitle = {Proceedings of the Genetic and Evolutionary Computation Conference Companion},
pages = {136–139},
series = {GECCO '22}
}

About

This is the official repository for the GECCO (Genetic and Evolutionary Computation Conference, 2022) paper "A Comparative Analysis on Genome Pleiotropy for Evolved Soft Robots". This repository hosts all the code necessary to replicate the experiments. This work was carried out at Ghent University IDLab-AIRO.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages