Code used to generate the paper "Heritability of morphological and behavioural traits in evolving robots"

Full path:

https://github.com/ci-group/revolve/tree/d683e64ce5ef98f0949138245aedd58e2b5d9423/experiments/karines_experiments/lsystem_cppn

Code used for the targeted locomotion in the Hardware Robots

Major Features and Improvements

• Refactored pygazebo
• Refactored population and most classes, completely different way of running the experiments
• Reintroduced the gzserver analyzer for self-collision checks
• Introduced docker image
• Many minor improvements

This is the first official release of the redisigned and improved Revolve framework.

Major Features and Improvements

• Code ported to Python 3
• Code ported to Gazebo 9
• Removed revolve-brain dependency
• Removed tol-revolve dependency
• Removed body-analyzer component
• Created revolve.py starting script
• Made it easier to debug the different components

Bug Fixes and Other Changes

• Terminal output is more complete and more responsive

Bug Introduced

• Running the brains on hardware has not been re-introduced

Thanks to our Contributors

Milan Jelisavcic, Matteo De Carlo

We are also grateful to all who filed issues or helped resolve them, asked and answered questions, and were part of inspiring discussions.

This release is ready for a single robot on-line evolution. Parts that were working for a multi-robot environment are broken and requires extra attention.

Major Features and Improvements

• Implemented rlpower algorithm
• Implemented two NEAT libraries
• Implemented Lamarckian evolution

Bug Fixes and Other Changes

• Run multiple experiments using script run-experiments.sh

Bug Introduced

• Not possible to run a multi-robot experiment with current setup

Thanks to our Contributors

Elte Hupkes, Milan Jelisavcic, Matteo De Carlo, Dmitry Egorov, Rafael Kiesel

We are also grateful to all who filed issues or helped resolve them, asked and answered questions, and were part of inspiring discussions.

This release is ready for a single robot on-line evolution. Parts that were working for a multi-robot environment are broken and requires extra attention.

Major Features and Improvements

• Implemented rlpower algorithm
• Implemented two NEAT libraries
• Implemented Lamarckian evolution

Bug Fixes and Other Changes

• Run multiple experiments using script run-experiments.sh

Bug Introduced

• Not possible to run a multi-robot experiment with current setup

Thanks to our Contributors

Elte Hupkes, Milan Jelisavcic, Matteo De Carlo, Dmitry Egorov, Rafael Kiesel

We are also grateful to all who filed issues or helped resolve them, asked and answered questions, and were part of inspiring discussions.