Skip to content
No description, website, or topics provided.
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.



Run our ROS kinetic environment including the workspace and projects

cd workspace_included

A container (robopaas/rosdocked-kinetic-workspace-included) will be pulled and started, it will have access to your X server.

You can try our projects within it, e.g., to run the robot navigation project:

roslaunch icclab_summit_xl irlab_sim_summit_xls_complete.launch

Or to run the grasping project:

roslaunch icclab_summit_xl irlab_sim_summit_xls_grasping.launch

For the grasping project you can test the python script to grasp the given object in the environment:

cd catkin_ws/src/icclab_summit_xl/scripts

Our default setup uses a Schunk gripper. You can simulate also a Robotiq gripper by setting a parameter:

roslaunch icclab_summit_xl irlab_sim_summit_xls_grasping.launch robotiq_gripper:=true
cd catkin_ws/src/icclab_summit_xl/scripts


If GUI-based apps don't work on your linux you'll have to allow the container to connect to your x-server:

(unsafe) before starting the container, on the host run:

xhost +

Longer story (Only if you need to rebuild / edit code)

Run ROS Kinetic / Ubuntu Trusty within Docker on Ubuntu Xenial or on any platform with a shared username, home directory, and X11.

This enables you to build and run a persistent ROS Indigo workspace as long as you can run Docker images.

Note that any changes made outside of your home directory from within the Docker environment will not persist. If you want to add additional binary packages without having to reinstall them each time, add them to the Dockerfile and rebuild.

For more info on Docker see here:

Rather than having devs build the entire image, we split the build logic in two steps:

  • in the main folder creates a user agnostic container image (robopaas/rosdocked-kinetic:latest)
  • personalized_image/ adds the current user to the image to mount home dir and access X server

Build (ICCLab image - optional)

This will create a common image for ICCLab (robopaas/rosdocked-kinetic:latest) still without your user/group ID and home directory.


Build personalized image

To add the current user to the image to be able to mount home dir and access X server, run:

cd personalized_image/

This will pull robopaas/rosdocked-kinetic:latest (hence no need to build it in the optional step)

Run personalized image

After building, this will run the docker image.


The image shares its network interface with the host, so you can run this in multiple terminals for multiple hooks into the docker environment.

Once in the container you should source the devel/setup.bash file in your (kinetic) catkin workspace from your home directory

Notes on what we added to the Dockerfile

  • Turtlebot install files
  • Move_it
  • Point Cloud library (manually built deb files)
  • fake Realsense camera packages (manually built deb files)

Notes on how we built deb files

Dockerfile workaround for realsense camera

COPY ros-kinetic-librealsense.postinst /ros-kinetic-librealsense.postinst COPY ros-kinetic-librealsense.control /ros-kinetic-librealsense.control RUN apt-get download ros-kinetic-librealsense &&
mkdir tmp_deb && cd tmp_deb &&
ar p ../ros-kinetic-librealsense_1.12.1-0xenial-20180809-140204-0800_amd64.deb control.tar.gz | tar -xz &&
cp ../ros-kinetic-librealsense.postinst postinst && cp ../ros-kinetic-librealsense.control control &&
cp ../ros-kinetic-librealsense_1.12.1-0xenial-20180809-140204-0800_amd64.deb ../ros-kinetic-librealsense_1.12.1icclab-0xenial-20180809-140204-0800_amd64.deb &&
tar czf control.tar.gz *[!z] &&
ar r ../ros-kinetic-librealsense_1.12.1
icclab-0xenial-20180809-140204-0800_amd64.deb control.tar.gz &&
cd .. && dpkg -i ros-kinetic-librealsense_1.12.1~icclab-0xenial-20180809-140204-0800_amd64.deb &&
rm -rf /tmp_deb #&& apt-get remove -y dkms && apt -o APT::Sandbox::User=root update

Dockerfile build of deb packages for python point cloud library (PCL)

PCL library (point cloud python -- avoid re-running, it takes forever!

RUN apt-get update -y && apt-get install -y build-essential devscripts dh-exec python-sphinx doxygen doxygen-latex RUN add-apt-repository --remove ppa:v-launchpad-jochen-sprickerhof-de/pcl -y &&
dget -u &&
cd pcl-1.7.2 && DEB_BUILD_OPTIONS=nodocs dpkg-buildpackage -j3 -r -uc -b RUN dpkg -i pcl.deb

You can’t perform that action at this time.