Roborobo is a fast and simple 2D mobile robot simulator loosely based on low-cost mobile robots such as khepera or epuck models. It is targeted for fast single and multi-robots simulation for evolutionary robotics and machine learning in multi-agent systems, collective and swarm robotics .
Roborobo combines speed of development and speed of execution. Roborobo can be programmed with python 3.x, with all the core functions written in C++ for super fast execution.
Roborobo version 4 is currently the only supported version. Current official release is (from: roborobo.cpp):
- gVersion = 20210321
- gCurrentBuildInfo = Shangri-La build
Main contributors
-
Nicolas Bredeche: main roborobo developper and project initiator (since 2009)
-
contact: nicolas.bredeche(at)sorbonne-universite.fr
-
Paul Ecoffet: pyRoborobo, the python interface to Roborobo (2020-2021)
-
Evert Haasdijk: properties management library (2010-2012)
Other contributors
- Jean-Marc Montanier, Berend Weel, Amine Boumaza, Andreas Steyven, Leo Cazenille, Theotime Grohens, and a few others!
If you use Roborobo in your work, please cite the following paper:
N. Bredeche, J.-M. Montanier, B. Weel, and E. Haasdijk. Roborobo! a fast robot simulator for swarm and collective robotics. CoRR, abs/1304.2888, 2013.
Link to the paper on Arxiv: http://arxiv.org/abs/1304.2888
Scientific papers that cite Roborobo: https://scholar.google.fr/scholar?cites=7785979290259259170
Roborobo basic dependencies are:
- a C++ compiler (GCC or CLANG)
- Python 3.x
Supported platforms:
- Linux-based
- MacOS X
Linux and MacOS installation instructions are described below. Other platforms are not officially supported, but Roborobo was previously shown to run on: MS Windows, Raspbian and Pandora.
Remark: if you get a lot of warnings during compilation, this is probably due to already installed pip packages shadowing the newly installed conda packages (e.g. with pybind). Work around for pyBind that may work: conda install -c conda-forge "pybind11>2.6". However the best way is to delete the pip packages and make a clean install of roborobo again_
Create a conda environment:
conda create --name roborobo numpy pybind11
conda activate roboroboInstall Python dependencies for Roborobo (numpy, pybind11, sphinx, ...)
conda install numpy setuptools
conda install -c conda-forge pybind11
conda install sphinx recommonmark sphinx_rtd_theme numpydoc
# or if you prefer to use pip instead of Conda:
# pip install -U pybind11
# pip install -U numpy
# pip install -U sphinxInstall C++ dependencies for Roborobo (Cmake, SDL2, boost and eigen):
sudo apt install git build-essential cmake
sudo apt-get install libsdl2-dev libsdl2-image-dev libboost-dev libeigen3-devGet your local copy of Roborobo:
git clone https://github.com/nekonaute/roborobo4.gitCompile and install Roborobo:
# conda activate roborobo (if not already activated)
cd roborobo4
python setup.py install --forceCheck the QUICK START section below for running a Roborobo example.
Create a conda environment:
conda create --name roborobo numpy pybind11
conda activate roboroboInstall Python dependencies for Roborobo (numpy, pybind11, sphinx, ...)
conda install numpy setuptools
conda install -c conda-forge pybind11
conda install sphinx recommonmark sphinx_rtd_theme numpydoc
# or if you prefer to use pip instead of Conda:
# pip install -U pybind11
# pip install -U numpy
# pip install -U sphinxInstall C++ dependencies for Roborobo (Cmake, SDL2, boost and eigen):
brew install cmake
brew install sdl2
brew install sdl2_image
brew install boost
brew install eigenGet your local copy of Roborobo:
git clone https://github.com/nekonaute/roborobo4.gitCompile and install Roborobo:
# conda activate roborobo (if not already activated)
cd roborobo4
python setup.py install --forceCheck the QUICK START section below for running a Roborobo example.
It is highly suggested to use the python interface to Roborobo, which we refer to as pyRoborobo. If you prefer to develop your project in C++, it also possible (check below). pyRoborobo is built as an interface to Roborobo, and though there is of course a cost to use Python instead of pure C++, we empirically consider it worth the ease of development in the context of academic research. For example, the Boids example runs at ~400 fps (pure C++) and ~200 fps (pyRoborobo) on a Macbook pro 13 (early 2019 model).
Roborobo (C++) and pyRoborobo (Python) both uses three important directories, that should be accessible from where your code (C++ binary or python script) is run.
- data contains image and resources for setting a roborobo environment
- config contains configuration files for running a roborobo environment
- logs will contain log files generated during a roborobo run
While running an example, type "h" when the focus is on the Roborobo window. Help tips will be displayed in the console.
Activate conda environment (if not done already):
conda activate roboroboCompile and install Roborobo (if not done already):
cd <your_roborobo_folder>
# python setup.py clean --all -- only if want to rebuild all from scratch
python setup.py install --force Run a pyRoborobo example:
cd <your_roborobo_folder>/pyRoborobo_dev/examples/
python tutorial.pyMany other examples are available in the pyRoborobo_dev/examples folder.
Build Roborobo's python API documentation:
# conda activate roborobo (if not already activated)
python setup.py build_sphinxThe pyRoborobo API documentation is now in build/sphinx/html/index.html
If you installed Roborobo for the first time, setup the directory structure for running Roborobo:
Setup the directory structure for both C++ and Python development (done only once):
# Roborobo C++ code
cd <your_roborobo_folder>/build
ln -s ../data
ln -s ../config
ln -s ../logsActivate conda environment (if not done already):
conda activate roboroboCompile and install Roborobo (to be done everytime you modify the C++ code:
cd <your_roborobo_folder>
# python setup.py clean --all -- only if want to rebuild all from scratch
python setup.py install --force Run a roborobo example:
cd <your_roborobo_folder>/build
./roborobo -l config/Tutorial.propertiesRoborobo (C++) examples can be found in the <your_roborobo_folder>/prj directory. Note that project selection is achieved from the configuration file (config/filename.properties)
- Check _OVERVIEW.TXT for a quick introduction.
- Check _FAQ.TXT for trouble shooting and frequently asked questions.
- Check the examples, and learn by doing.
Thank you for using Roborobo!