This is an example of orocos component that connects to the FRI:
Run the KRL script on the robot, then the orocos example script.
This component depends on the ROS package kuka_robot_hardware
Go to the ros_workspace directory, and add the package:
cd ros_workspace
wstool set RTEKukaFRIClient --git https://github.com/XDE-ISIR/RTEKukaFRIClient.git
wstool update RTEKukaFRIClient
First launch roscore: roscore
then in another terminal launch the openNi node: roslaunch openni_launch openni.launch
Setup the robot in GravityComp mode and runfri_open, runfri_start
Run the initialization script for the calibration rosrun ocl deployer-gnulinux -s Demo_disposvel01calibration.ops
Run the calibration program and follow the instructions: roslaunch kuka_kinect_calibration calibrate_kuka.launch
publish the /world to /camera_link transformation returned by the calibration program, example: rosrun tf static_transform_publisher 0.755449 -1.47254 1.16996 .166774 0.0701399 0.931835 0.314563 /world /camera_link 100
verify by launching rqt: rqt
Changed the fixed frame to /camera_link Add a Tf object Add a PointCloud2 object and change the topic to /camera/depth_registered/points to visualize the kinect image Check that the robot segments correspond to the robot image
Run the human tracker: rosrun tracker tracker
Setup the robot in PositionControl mode, and run the script krl.src
Run the second demo script type: rosrun ocl deployer-gnulinux -s Demo_disposvel02start.ops
Add a path object in rqt and select the topic /distances/path to visualize minimum distance between the human limbs and the robot links