Make sure it was Indigo and Ubuntu 14.04 :
printenv | grep ROS
lsb_release –a
Do not forget to update to latest software update for ROS and Ubuntu
sudo apt-get update
sudo apt-get dist-upgrade
Turtlebot arm: The turtlebot arm packages provides bringup, description and utilities for using the PhantomX Pincher arm. Installation from source:
cd ~/ros/indigo/catkin_ws/src
git clone https://github.com/IOJVision/Turtlebot_arm.git
cd .. && catkin_make
ArbotiX Ros: ArbotiX ROS drivers let the ROS interface with the ArbotiX board on the PhantomX Pincher Installation from source:
cd ~/ros/indigo/catkin_ws/src
git clone https://github.com/IOJVision/arbotix_ros.git
cd .. && catkin_make
MoveIt!: MoveIt! let you plan the arm movement and visualize the arm in RViz Installation from deb:
sudo apt-get install ros-indigo-moveit
To edit or view the arm architecture:
roscd turtlebot_arm/turtlebot_arm_bringup/config
gedit pincher_arm.yaml
The arm architecture is described in yaml file. USB port, number of joints, limits...
source: pincher_arm.yaml
port: /dev/ttyUSB0
read_rate: 15
write_rate: 25
joints: {
arm_shoulder_pan_joint: {id: 1, neutral: 512, max_angle: 140, min_angle: -140, max_speed: 90, type: dynamixel},
arm_shoulder_lift_joint: {id: 2, max_angle: 126, min_angle: -119, max_speed: 90, type: dynamixel},
arm_elbow_flex_joint: {id: 3, max_angle: 136, min_angle: -139, max_speed: 90, type: dynamixel},
arm_wrist_flex_joint: {id: 4, max_angle: 96, min_angle: -98, max_speed: 90, type: dynamixel},
gripper_joint: {id: 5, max_angle: 0, min_angle: -145, max_speed: 90, type: prismatic, radius: .0078, connector: .024, offset: .016}
}
controllers: {
arm_controller: {type: follow_controller, joints: [arm_shoulder_pan_joint, arm_shoulder_lift_joint, arm_elbow_flex_joint, arm_wrist_flex_joint], action_name: arm_controller/follow_joint_trajectory, onboard: False }
}
- Make sure the FTDI cable is already plugged in and all the servo cable is already attached correctly
- As the turtlebot arm package port is already set to “/dev/ttyUSB0”, make sure you got the right port or you can change all the port setup at the .yaml file inside the turtlebot arm bringup package.
- After all the check is done, start your terminal.
- Start to bringup the arm first, roslaunch turtlebot_arm_bringup arm.launch
- Then start the controller, rosrun arbotix_python arbotix_gui
- After the bringup and the controller is successfully run, start to play with the arm. You can turn on or off the servo in the arbotix_gui controller.
SOLUTION:
sudo adduser <the user you want to add> dialout
sudo reboot
Simulation:
-
Make sure RViz is installed in your computer.
-
Go to the turtlebot arm moveit launch file then change the argument of “sim” to “true”.
roscd turtlebot_arm_moveit_config/launch gedit turtlebot_arm_moveit.launch -
Then run the turtlebot arm moveit launch file
roslaunch turtlebot_arm_moveit_config turtlebot_arm_moveit.launch -
Or by change by the launch command, run the turtlebot arm moveit launch file
roslaunch turtlebot_arm_moveit_config turtlebot_arm_moveit.launch sim:=true --screen
Real arm:
-
Plug in the arm usb to the computer
-
Go to the turtlebot arm moveit launch file then change the argument of “sim” to “false”. follow as above step..
-
Run the turtlebot arm bringup
roslaunch turtlebot_arm_bringup arm.launch without the AbotiX ROS roslaunch turtlebot_arm_bringup armcontrol.launch with the ArbotiX ROS -
Then as usual run the turtlebot arm moveit launch file
roslaunch turtlebot_arm_moveit_config turtlebot_arm_moveit.launch
-
Start up the arm and the camera for testing
roslaunch turtlebot_arm_bringup arm.launch for the arm roslaunch freenect_launch freenect.launch for the kinect -
Check all to make sure it is functioning then terminate it.
-
To start calibration,
__Make sure to have printed Calibration checkerboard (7x6 27mm) __: click here to download the checkerboard: checkerboard
The complete checkerboard must be in the kinect field of view :
Then, manually try to move the arm to make sure it can reach all the point in the checkerboard.:
Additional info, the checkerboard must be on the same level of the arm base. :
-
After all the test and set up has been done correctly, run the calibration program
roslaunch turtlebot_arm_kinect_calibration calibrate.launch In that launch file is already include the camera launch and arm bringup also the servo controller. -
Firstly, it will pop out the image view of the calibration pattern on the checkerboard if it’s able to detect it and the arm servo controller.
-
The trick is the image by the calibration image view is just the picture of the pattern and it does not show the live picture of the movement of the arm, to view the live movement of the arm, can use rosrun image_view with the kinect camera topic or by using the RViz.
-
To move the arm to all the necessary point is very easy if we do by manually and less time taken for the calibration process. Start with uncheck all the servo, to let the motor power off so that we can move it manually but make sure the gripper is widely open and use the right section of the gripper to hit the 4-calibration point.
-
Start with move it to the first point then press enter, continue until the 4th point.
First Point:
Second Point:
Third Point:
Fourth Point:
-
Until all the point reaches, in the terminal will process and it will print out the result like this:-
Resulting transform (camera frame -> fixed frame): 0.999483 -0.028142 0.0155747 -0.184333 -0.0182255 -0.894504 -0.446689 0.314301 0.0265024 0.446174 -0.894554 0.951596 0 0 0 1 Resulting transform (fixed frame -> camera frame): 0.0265023 0.446174 -0.894554 0.951596 -0.999483 0.0281421 -0.0155746 0.139333 0.0182256 0.894504 0.446689 -0.314301 4.59163e-41 0 0 1 Static transform publisher (use for external kinect): rosrun tf static_transform_publisher x y z qx qy qz qw frame_id child_frame_id period_in_ms rosrun tf static_transform_publisher 0.119769 -0.147356 0.993819 -0.371373 0.372475 0.589925 0.612645 /base_link /camera_link 100 URDF output (use for kinect on robot): <?xml version="1.0"?> <robot> <property name="turtlebot_calib_cam_x" value="0.119769" /> <property name="turtlebot_calib_cam_y" value="-0.147356" /> <property name="turtlebot_calib_cam_z" value="0.993819" /> <property name="turtlebot_calib_cam_rr" value="-0.0348527" /> <property name="turtlebot_calib_cam_rp" value="1.10743" /> <property name="turtlebot_calib_cam_ry" value="1.51147" /> <property name="turtlebot_kinect_frame_name" value="base_link" /> </robot> -
Run the static transform publisher output and As long as this command runs, the static_transform_publisher will publish the transform between the arm frame and the kinect frame. If you move the physical camera, you will need to recalibrate again
rosrun tf static_transform_publisher 0.366333 0.227789 0.984579 0.436052 0.44573 -0.573806 0.530971 /base_link /camera_link 100
-
Start the MoveIt from turtlebot arm package in the simulation mode(by default, the simulation is false)
roslaunch turtlebot_arm_moveit_config turtlebot_arm_moveit.launch sim:=true –-screen -
After the commad has been launch, you will see the PhantomX Pincher robot arm model in the RViz. If it show other robot arm model, please change the environment variable from turtlebot_arm1 to pincher.
-
After that start the pick and place demo by running the below line in the new terminal.
rosrun turtlebot_arm_moveit_demos pick_and_place.py
-
For the real arm, we need to use calibration point as we did in above then use the same set up as that cause we do not want to calibrate again if we move the camera or change the set up.
-
Firstly, we need to have a cube type of object with dimension of 2 cm or 2.5 cm because our gripper is only capable to grab until 3 cm. We can change the dimension of the box inside the launch file and also if our table is not aligned to the base of the arm, we can set that also inside the launch file.
<launch> <!-- ************* Arm bringup stuff ************* --> <include file="$(find turtlebot_arm_bringup)/launch/arm.launch" /> <node name="arbotix_gui" pkg="arbotix_python" type="arbotix_gui" output="screen"/> <!-- ************* Moveit config stuff ************* --> <!-- Load the URDF, SRDF and other .yaml configuration files on the param server --> <include file="$(find turtlebot_arm_moveit_config)/launch/planning_context.launch"> <arg name="load_robot_description" value="true"/> </include> <!-- Run the main MoveIt executable to provide move groups --> <include file="$(find turtlebot_arm_moveit_config)/launch/move_group.launch"> <arg name="allow_trajectory_execution" value="true"/> <arg name="fake_execution" value="false"/> <arg name="info" value="true"/> </include> <!-- ************* Kinect bringup stuff ************* --> <include file="$(find freenect_launch)/launch/freenect.launch" />s <!-- ************* Block manipulation stuff ************* --> <include file="$(find turtlebot_arm_block_manipulation)/launch/block_manipulation.launch" /> <node name="block_manipulation_demo" pkg="turtlebot_arm_block_manipulation" type="block_manipulation_demo" output="screen" > <param name="arm_link" value="/arm_base_link" /> <param name="gripper_open" value="0.1" /> <param name="gripper_closed" value="0.001" /> <param name="z_up" value="0.1" /> <param name="table_height" value="-0.03" /> <param name="block_size" value="0.020" /> in here we have to change the size of the object <!--param name="target_x" value="" /--> <!--param name="target_y" value="" /--> </node> <node name="rviz" pkg="rviz" type="rviz" args="-d $(find turtlebot_arm_block_manipulation)/demo/block_manipulation_demo.rviz" /> <node pkg="tf" type="static_transform_publisher" name="static_transform_publisher" args="0.552972 0.0311763 1.01794 -0.485351 0.0823443 0.864462 0.101775 /base_link /camera_link 100"/> THIS IS THE CALIBRATION POINT WE INCLUDED IN THE LAUNCH FILE </launch> -
For now, the code is only capable to detect green cube only.
-
To start the block detection, launch
roslaunch turtlebot_arm_block_manipulation block_manip_complete.launchIf our calibration is already good then we do not want to change the position, we can add permanently in our launch file.
-
Inside the launch file that we launch in step 4, it is included with the arm bringup, camera bringup and also rviz moveit bringup.
-
After all of the necessary application launch, run this command to start object detetction and demonstration of pick and place
rostopic pub -r 1000 /relay/robot_status2 std_msgs/String "vsdone" -
Demonstration video for: 2cm cube 2.5cm cube
-
Before starting, make sure you have the rbx1 vision package from pirobot
cd ~/ros/indigo/catkin_ws/src git clone https://github.com/IOJVision/rbx1.git cd .. && catkin_make -
First, run this command of bringup the arm, moveit rviz and kinect
roslaunch turtlebot_arm_bringup arm_moveit.launch -
Then run the face detector
roslaunch rbx1_vision face_tracker2.launch -
Lastly, run the script
rosrun turtlebot_arm_moveit_demos head_tracker.py -
When there is face detected, the arm will pose a waving pose and when there is no face, the arm will be in resting.