Introduce to ROS

• Introduce to ROS
  • Simulate a Denso robot in ROS environment
    • Install ROS and package
    • Pull Denso model from github
  • Motion planning with MoveIt framework
  • Add a pattern object to Gazebo
  • Contorl real Denso robot
    • With Teaching Pendant
    • With Wincap III
    • With C Sharp
    • With ROS

Simulate a Denso robot in ROS environment

Install ROS and package

• OS requirement: Ubuntu 16.04 LTS.

• Make sure your python2 enviroment up to date, check the version by run the following command in terminal

  python2 --version
  pip2 --version
  

  Update your python2:

  sudo apt-get update
  sudo apt-get install -y python
  sudo apt-get install -y python-pip
  sudo pip2 install --upgrade pip
  

• Install ROS Kinetic and MoveIt package, save the content to the file install_ros.sh, and run bash install_ros.sh in terminal.

  #!/bin/bash
  
  sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu $(lsb_release -sc) main" > /etc/apt/sources.list.d/ros-latest.list'
  sudo apt-get update
  sudo apt-get install -y --allow-unauthenticated ros-kinetic-desktop-full
  apt-cache search ros-kinetic
  
  echo "" >> ~/.bashrc
  echo "# ROS" >> ~/.bashrc
  echo "export EDITOR='code'" >> ~/.bashrc
  echo "source /opt/ros/kinetic/setup.bash" >> ~/.bashrc
  source ~/.bashrc
  
  sudo rosdep init
  rosdep update
  
  sudo apt-get install -y --allow-unauthenticated ros-kinetic-moveit
  sudo apt-get install -y --allow-unauthenticated ros-kinetic-denso-robot-ros
  
  source /opt/ros/kinetic/setup.bash
  
  sudo mkdir -p /data/denso_ws/src/ros_home/img
  sudo chown -R $USER:$USER /data
  cd /data/denso_ws/
  catkin_make
  echo "source /data/denso_ws/devel/setup.bash" >> ~/.bashrc
  echo "export ROS_HOME=/data/denso_ws/src/ros_home" >> ~/.bashrc
  
  source ~/.bashrc
  

Pull Denso model from github

• Close termial and restart it
• Now we need to remove the original denso model and pull the new one from github

  cd /opt/ros/kinetic/share    # Change to ROS directory
  ls denso*                    # list all package of denso
  sudo cp -rf denso* ~/Desktop # Backup to Desktop
  sudo rm -rf denso*           # Delete them!
  

• Check is there a directory in the path /data/denso_ws. if not, run those command to build a catkin workspace

  sudo mkdir -p /data/denso_ws/src/ros_home/img
  sudo chown -R $USER:$USER /data
  cd /data/denso_ws/
  catkin_make
  echo "source /data/denso_ws/devel/setup.bash" >> ~/.bashrc
  echo "export ROS_HOME=/data/denso_ws/src/ros_home" >> ~/.bashrc
  

• Pull the Denso ROS model by running:

  cd /data/denso_ws/src
  sudo apt-get install -y git
  git clone https://github.com/solab-ntu/denso_robot_ros.git
  

• Test it

  roslaunch denso_robot_bringup vs060_bringup.launch
  

Motion planning with MoveIt framework

There are three interface to control robot with MoveIt:

1. Rivz GUI Launch Denso vs060 robot, and you should open Gazebo and Rviz window

   roslaunch denso_robot_bringup vs060_bringup.launch
   

   There is a moveit GUI plugin in Rviz interface, just drag the end effector and excute robot as shown in the GIF

2. Commander Scripting Use shell-like command interact with robot, first run the command in the new terminal windows,

   roslaunch moveit_cmdline_commander moveit_cmdline_commander.launch
   

   The command below will start a command line interface tool that allows you to connect to a running instance of the move_group node. The first command you should type is:

   use arm
   

   This will connect to the move_group node for the group name you specified (in the Denso, for instance, you could connect to arm). You can now execute commands on that group. This command, current, will show you the current state of your group:

   current
   

   To record that state under a specific name you can simply type:

   rec goal
   

   This will remember the current joint values of the robot group under the name goal. Matlab-like syntax is available for modifying joint values. We then modify the first joint of goal to 0.2. You may need to use a different value instead of 0.2 (it needs to be within your allowed bounds and not cause collisions). The go command plans a motion and executes it.

   To get the robot to move, you could type, for example:

   goal[0] = 0.2
   go goal
   

   Instead of calling go you could also type:

   goal[0] = 0.2
   goal[1] = 0.2
   plan goal
   execute
   

   This is slightly inefficient, but the advantage is that the plan command allows you to visualize the computed motion plan in RViz before you actually issue the execute command.

   For a list of supported commands, you can type help. To exit the moveit_commander interface you can type quit.

3. Python interface

   Run the command in the new terminal windows,

   roslaunch moveit_python_commander moveit_python_commander.launch
   

   In RViz, we should be able to see the following:

   Press in the shell terminal where you ran the command in between each step

   1. The robot plans and moves its arm to the joint goal.
   2. The robot plans a path to a pose goal.
   3. The robot plans a Cartesian path.
   4. The robot executes the Cartesian path plan.

Add a pattern object to Gazebo

1. Generate pattern texture from VISIO or OpenCV. Here is a example

1. Build pattern object(*.dae) with Blender Install Blender from website and here is tutorial

   Finally, you should generate a pattern object in *.dae format, now we should add this object to the Denso Gazebo urdf model. (Make sure there is pattern.dae file and pattern.png texture in denso_robot_descriptions folder)

   # vs060.urdf locate in denso_robot_descriptions/vs060_description
   
   <robot name="vs060">
   ... vs060 robot description ...
   
   <!-- Pattern -->
   <link name="pattern">
       <visual>
           <origin rpy="0 0 0" xyz="0 0 0" />
           <geometry>
               <mesh filename="package://denso_robot_descriptions/vs060_description/pattern.dae" scale="1 1 1" />
           </geometry>
       </visual>
       <collision>
           <origin rpy="0 0 0" xyz="0 0 0" />
           <geometry>
               <box size="0.22 0.18 0.002"/> 
           </geometry>
       </collision>
       <inertial>
           <mass value="1" />
           <origin rpy="0.000000 0.000000 0.000000" xyz="0.000000 0.000000 0.000000" />
           <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1" />
       </inertial>
   </link>
   
   <joint name="joint_pattern" type="fixed">
       <parent link="world" />
       <child link="pattern" />
       <origin rpy="0.000000 0.000000 1.5707963" xyz="0.300000 0.000000 0.001" />
   </joint>
   
   <!-- Add emissive light in pattern self -->
   <gazebo reference="pattern">
       <visual>
           <material>
               <emissive>0.5 0.5 0.5 0.5</emissive>
               <shader type='vertex'>
                   <normal_map>__default__</normal_map>
               </shader>
           </material>
       </visual>
   </gazebo>
   </robot>
   

Contorl real Denso robot

With Teaching Pendant

Basic operation

With Wincap III

1. Connect PC and RC8 control with ethernet cable. Then, change local IP in same LAN Network with RC8, 192.168.0.2 for example. Programming tutorial:

With C-Sharp

1. Install ORiN2 SDK and Visual Studio
2. Clone the C-Sharp sample code from here

With ROS

1. Setting up the bCap server in Denso robot, please see here.
2. Generate URDF model for VP6242 and an associated MoveIt configuration package as a standard component, please see here
3. Connect your linux computer and RC8 control with ethernet cable. Then, change local IP in same LAN Network with RC8, 192.168.0.2 for example. Control robot by MoveIt framework, simply run

   roslaunch denso_robot_bringup vs6242_bringup.launch sim:=false ip_address:=192.168.0.1