Skip to content
Paper code for `Data-efficient Learning of Morphology and Controller for a Microrobot` (ICRA 2019)
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.


Type Name Latest commit message Commit time
Failed to load latest commit information.

Learning Morphology and Control for a Microrobot

This is the repository for the paper "Learning Morphology and Control for a Microrobot". More information can be found on our website here. Included are demos for running the experiments laid out in the paper.


Most of the pip dependencies can be installed with pip install -r requirements.txt. First activate the venv with source venv/bin/activate, then install GPy with

git clone
pip install -e GPy

Due to the complexity of managing different python versions and dependencies between different modules, we recommend running this repository on the associated docker image.

Running Experiments

Experiments can be run either locally or via Docker.

Running locally

First, start V-REP running locally. Then use to call the relevant experiment, for example:

python hpcbbo --init_uc=5 --init_cn=5 --uc_runs_per_cn=5 \
                      --batch_size=5 --total=5 --obj_f=1 --contextual


python random --init_uc=5 --total=5 --obj_f=1

Call python -h for additional help

Running on Docker

Instructions for setting up Docker are included in the related repository.

Simulator Setup

Before running any of the experiments, make sure V-REP is open (see the V-REP documentation for troubleshooting issues with installation/booting). Scenes are automatically loaded and can be found in scenes/. The default simulator settings should work fine, but check that the following settings are correct:

  • Physics engine: Bullet 2.78
  • Time step: 50 ms


Should you find this code useful, please support us by citing our paper:

T. Liao, G. Wang, B. Yang, R. Lee, S. Levine, K. Pister, R. Calandra. 
Data-efficient Learning of Morphology and Controller for a Microrobot.
In IEEE Int. Conf on Robotics and Automatation, ICRA '19, Montreal, Canada, May


  title                 = {Data-efficient Learning of Morphology and Controller for a Microrobot},
  author                = {Liao, Thomas and Wang, Grant and Yang, Brian and Lee, Rene and Pister, Kristofer and Levine, Sergey and Calandra, Roberto},
  booktitle             = {2019 IEEE International Conference on Robotics and Automation},
  year                  = {2019},
  url                   = {}
You can’t perform that action at this time.