robot_state_publisher : process has died every time when start up the PR2 robot
On the PR2 robot, every time the robot is booted, we have to roslaunch /etc/ros/robot.launch to start up the essential nodes.
However, the robot_state_publisher node died every time when I roslaunch /etc/ros/robot.launch. Here's the output message:
[head_traj_controller/point_head_action-41] process has died [pid 13035, exit code -11, cmd /opt/ros/groovy/stacks/pr2_controllers/pr2_head_action/bin/pr2_head_action /diagnostics:=/diagnostics /diagnostics_agg:=/diagnostics_agg __name:=point_head_action __log:=/home/pr2admin/.ros/log/ac1de78a-26ae-11e3-9f32-001517ebc33d/head_traj_controller-point_head_action-41.log].
log file: /home/pr2admin/.ros/log/ac1de78a-26ae-11e3-9f32-001517ebc33d/head_traj_controller-point_head_action-41*.log
I've done some basic testings:
1.Suppose that there are some problems in URDF.
Since the robot_state_publisher need the robot URDF parameter in the parameter server and build up the TF tree. I guess if there are some problems in the URDF that cause the process to die every time. Therefore I test the uncalibrated URDF and the original URDF generated from the xacro, but The process still died every time it was launched.
2.Suppose that there are some problems in the robot_state_publisher executable file. I first upload the robot_state_publisher executable file in my computer, which works fine when running gazebo simulation, onto PR2. But the process still died. Then I try to build the package from source, but I still get the same problem.
3.I then run the node singly with required parameters such as the URDF and repeat the procedure above, but the problem is still there. But all is well when I run it on my own computer. The problem only happen on the PR2 computer.
What might the problem be? Thanks for any advise.
