Skip to content

Repository for the compliant humanoid robot COMAN from the Italian Institute of Technology.

License

Notifications You must be signed in to change notification settings

cogimon/iit-coman-ros-pkg

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

iit-coman-ros-pkg

Repository for the compliant humanoid robot COMAN from the Italian Institute of Technology contains a set of ROS packages. The updated packages are coman_urdf, coman_gazebo and coman_srdf while the others are no longer updated.

This repository can be used with ROS with a limited set of functionalities.

#Package list:

coman_urdf:

contains the urdf.xacro description of COMAN as well as files that are needed in the simulation in GAZEBO. Every time one of these files are changed you have to use the script inside the script/ folder to generate. The script generates files that are used in simulation as well as files that can be used inside kinematics/dynamics libraries (coman.urdf). Running this script requires ROS and some python dependencies. You can just install a precompiled version of the urdf and sdf, though. Instructions are in the coman_gazebo section of this readme.

After installing ros (tested on ros hydro), make sure the dependency on BeautifulSoup4 is satisfied before using the script:

sudo easy_install beautifulsoup4

If you have problems with this command maybe you have to install python or the package easy_install. After this, you have to be sure that coman_urdf is inside your ros path. Either put it in your ros package path or do something like this:

export ROS_PACKAGE_PATH=${ROS_PACKAGE_PATH}:/home/your_name/projects/walkman/iit-coman-ros-pkg

The script has to be called in script/ folder:

./create_urdf_and_sdf.sh

It generates the coman urdf and srdf descriptions and sdf model as specified in coman_config.urdf.xacro.

After running the script you can check the results directly by running

roslaunch coman_urdf coman_slider.launch
rosrun rviz rviz

or directly in simulation.

coman_gazebo:

contains files needed for the simulation as well as a launch file to start the simulation. The coman_gazebo folder includes precompiled sdf and urdf files for coman, as well as many initialization files for the setting of the control parameters for the simulated robot (YARP VERSION). To install these models files, you should first remove the version you already have installed (make sure to make a backup if you manually modified the configuration files)

rm -r ~/.gazebo/model/coman_urdf

And to install the new version, from the coman_gazebo folder, run

mkdir ~/.gazebo/models/coman_urdf
cp -R sdf/* ~/.gazebo/models/coman_urdf/

coman_srdf:

contains the kinematic description of COMAN intended as number of kinematich chains and joints included in each chain. These file is automatically generated by the script considering the coman_config.urdf.xacro.

coman_control (NO LONGER MANTAINED):

contains the files needed to control the robot. For now the available controller is the effort_controllers/JointPositionController. To start the controller:

roslaunch coman_control coman_control.launch

It is also implemented a node that permits to send trajectory to the robot that were previously recorded in a file, to run this node:

rosrun coman_control send_trajectories

in another terminal runs:

rosrun rqt_reconfigure rqt_reconfigure

remember to run this command in the same folder were is the traj.dat file. The format of this file can be found in coman_control/src.

coman_state_publisher (NO LONGER MANTAINED):

contains a node that permits to publish inside ROS the state of the real robot. Communication with COMAN occurs through xddp_sockets. This package has an external dependency against Robolli libraries. We are working to make it independent.

coman_msgs (NO LONGER MANTAINED):

contains msgs that are used in coman_state_publisher.

About

Repository for the compliant humanoid robot COMAN from the Italian Institute of Technology.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Shell 48.7%
  • CMake 25.7%
  • Python 25.2%
  • Makefile 0.4%