by RoboLab (Universidad de Extremadura), Aston University, ISIS (Universidad de Málaga) and many other collaborators from the Google Summer of Code program.

RoboComp is an open-source Robotics framework providing the tools to create and modify software components that communicate through public interfaces. Components may require, subscribe, implement or publish interfaces in a seamless way. Building new components is done using two domain-specific languages, IDSL and CDSL. With IDSL you define an interface and with CDSL you specify how the component will communicate with the world. With this information, a code generator creates C++ and/or Python sources, based on CMake, that compile and execute flawlessly. When some of these features have to be changed, the component can be easily regenerated and all the user-specific code is preserved thanks to a simple inheritance mechanism.

If you already have RoboComp installed, jump to tutorials to start coding!

⚠️ If you want to contribute with something new to Robocomp, please do it on the development branch.

• Installation from source
  • Requirements
  • Installation
• Testing the installation using the RCIS robotics simulator
  • Installing some RoboLab's components from GitHub
  • Connecting a JoyStick (if no JoyStick available skip to the next section)
  • Using the keyboard as a JoyStick
• Testing the installation using the Coppelia Simulator
• Next steps

Table of contents generated with markdown-toc

Tested in Ubuntu 18.04, 19.10, 20.04 and 20.10.
Note: RoboComp is not compatible with Ubuntu 16.04. RoboComp needs to be compiled using C++11. Ice libraries with C++11 support are only available for zeroc-ice 3.7 and the packages for this version are only available since Ubuntu 18.04.

Note: If you have installed Anaconda in your system. Then you need to change the python from anaconda to default.

Make sure you have installed the following packages from the Ubuntu repository (to copy and paste in a terminal use zash):

sudo DEBIAN_FRONTEND=noninteractive apt-get update && sudo DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends  \
      ca-certificates \
      cmake \
      curl \
      freeglut3-dev \
      g++ \
      gcc-multilib \
      git \
      git-annex \
      libboost-dev \
      libboost-system-dev \
      libboost-thread-dev \
      libgsl-dev \
      libopenscenegraph-dev \
      libqt5xmlpatterns5-dev \
      libqt5opengl5-dev \
      libzeroc-icestorm3.7 \
      make \
      python3 \
      python3-pip \
      python3-distutils \
      python3-prompt-toolkit \
      python3-pyparsing \
      python3-setuptools \
      python3-termcolor \
      sudo \
      zeroc-ice-all-dev \
      zeroc-icebox 

sudo pip3 install pyside2 rich

It is recommendable to install the following packages::

sudo apt-get install yakuake qttools5-dev-tools qt5-assistant

Note: One of the main tools of Robocomp, robocompdsl is using pyparsing and the current code doesn't work with 2.4 version of this library. With the previous commands, we are installing the 2.2 version (python-pyparsing=2.2.0+dfsg1-2). If you have a more recent version of pyparsing installed with apt or pip we recommend you to uninstall it and install the 2.2 version. You can check your current version of pyparsing with this command:

python3 -c "import pyparsing; print(pyparsing.__version__)"

cd to your home directory (you are probably in it already) and type:

git clone https://github.com/robocomp/robocomp.git

Now we will create a symbolic link so RoboComp can find everything. You will have to enter your password:

sudo ln -s ~ /home/robocomp

(the ~ is in Alt-4)

Edit your ~/.bashrc file

gedit ~/.bashrc

Add these lines at the end:

export ROBOCOMP=~/robocomp
export PATH=$PATH:/opt/robocomp/bin

make bash process the modified file by typing:

source ~/.bashrc

Done! Now let's compile and install the whole thing:

sudo [ -d /opt/robocomp ] && rm -r /opt/robocomp
cd robocomp
mkdir build
cd build
cmake ..
make
sudo env "PATH=$PATH" make install

If you want to compile Robocomp with support for FCL, follow the instructions in the Robocomp with FCL tutorial.

The RoboComp's core libraries and simulator should now be compiled and installed in /opt/robocomp.

Done! Now let's have some fun.

We will first fetch some meshes and textures used by the simulator (it will take a while):

cd ~/robocomp
git annex get .

Now let's run the simulator.

cd ~/robocomp/files/innermodel
rcis simpleworld.xml

You can also use the default innermodel/simpleworld.xml anywhere if you have set the ROBOCOMP environment variable.

rcis

Congratulations! RCIS should be up and running with a simple robot endowed with a laser and an RGBD camera, moving on a wooden floor. Don't forget to turn around the floor to see the robot from above.

The software of the robots using RoboComp is composed of different software components working together, communicating among them. What we just installed is just the core of RoboComp (the simulator, a component generator, and some libraries). To have other features like joystick control we have to run additional software components available from other repositories, for example, robocomp-robolab:

cd ~/robocomp/components
git clone https://github.com/robocomp/robocomp-robolab.git

The RoboLab's set of basic robotics components are now dowloaded. You can see them in ~/robocomp/components/robocomp-robolab/components

If you have a joystick around, connect it to the USB port and:

cd ~/robocomp/components/robocomp-robolab/components/hardware/external_control/joystickComp
cmake .
make
cd bin
sudo addgroup your-user dialout   // If you find permissions issues in Ubuntu
check the config file in the component's etc folder and male sure that the port matches the DifferentialRobot endpoint in     RCIS.
bin/joystick etc/config

Your joystick should be now running. It will make the robot advance and turn at your will. If it does not work, check where the joystick device file has been created (e.g., /dev/input/js0). If it is not /dev/input/js0, edit ~/robocomp/components/robocomp-robolab/components/hardware/external_control/joystickComp/etc/config change it accordingly and restart. Note that you might want to save the config file to the component's home directory so it does not interfere with future GitHub updates.

If you don't have a JoyStick install this component,

cd ~/robocomp/components/robocomp-robolab/components/hardware/external_control/keyboardrobotcontroller
cmake .
make
src/keyboardrobotcontroller.py etc/config

and use the arrow keys to navigate the robot, the space bar to stop it and 'q' to exit.

Note 1: You must have your simulator running in a terminal and only then you can run a component in another terminal. You will get an error message if you run the above component without having RCIS already running.

Note 2: If you have anaconda installed (for python 3), It is recommended to uninstall anaconda first and then install robocomp. (It is only applicable if you have faced errors while running above commands.)

We are now moving to more advanced robotics simulators that can reduce the gap between simulation and deployment. Our first choice now is CoppeliaSim because it offers a scene editor that can be used during a running simulation, you can "hang" and modify Lua code from the scene nodes in no time, you can choose among 4 physics engines and, thanks to the PyRep library, we have a fast access to almost eveything running in the simulator.

To connect RoboComp and CoppeliaSim we use bridges that are Python components that include PyRep and implement/publish the required RoboComp interfaces. So far we have implemented three bridges that are located in this repo. These bridges interface scenes with Viriato, a simpler world using Viriato's omni base and an empty world with a Pioneer 2AT differential robot. The corresponding Coppelia .ttt files can be found here.

To test RoboComp with Coppelia you need to:

• Install Coppelia Robotics and Pyrep. There are detailed instructions here
• Run the bridge, i.e. omniPyrep.py and see that Coppelia starts Ok.
• Connect your new component to the ports offered in omniPyrep.py or
• Connect a joystick or XBox pad to omniRep.py using this component

You can find more tutorials on RoboComp in tutorials

Drop comments and ask questions in:

• https://gitter.im/robocomp

Please, report any bugs with the github issue system: Robocomp Issues

If you have any suggestions to improve the repository, like features or tutorials, please contact us on or create a feature request here.