Robot arm falls over on spawn (Gazebo)
I'm using ROS Melodic with the denso_robot_ros package ( https://github.com/DENSORobot/denso_r... ).
I'm currently trying to adapt my own arm to utilize the package. The package comes with a functioning example using the VS060 model, whereas I'm attempting to use the VS6577 A-B model with a custom end effector. I have created my own URDF (in xacro format) by using data obtained from the VS6577 controller for the arm and the custom end effector files.
When running the VS060 example, the arm will spawn in both Gazebo and RViz, with RViz providing the ability to control the arm by moving the position indicator located at the tip. When executing a trajectory in RViz, the model in Gazebo will make movements corresponding to the trajectory.
When running the VS6577 arm, the arm will spawn in both Gazebo in RViz, however it will flop over in both and fail to respond to commands. The arm in RViz will mimic the movement of the arm in Gazebo and doesn't exert any force to correct itself/stand up. The RViz VS6577 will still display the clickable indicator on the end of the robot arm that allows me to move it to a hypothetical position that it can then successfully plan to move it. Once I decide to execute the trajectory however, the arm doesn't make any attempt at movement while the terminal indicates that it has successfully moved.
[ INFO] [1558814233.787314221, 39.284000000]: Combined planning and execution request received for MoveGroup action. Forwarding to planning and execution pipeline.
[ INFO] [1558814233.787400526, 39.284000000]: Planning attempt 1 of at most 1
[ INFO] [1558814233.787467946, 39.284000000]: Starting state is just outside bounds (joint 'joint_2'). Assuming within bounds.
[ INFO] [1558814233.787630107, 39.284000000]: Found a contact between 'EEF_robot_magnet' (type 'Robot link') and 'J4' (type 'Robot link'), which constitutes a collision. Contact information is not stored.
[ INFO] [1558814233.787663090, 39.284000000]: Collision checking is considered complete (collision was found and 0 contacts are stored)
[ INFO] [1558814233.787701964, 39.284000000]: Start state appears to be in collision with respect to group arm
[ INFO] [1558814233.789812451, 39.287000000]: Found a valid state near the start state at distance 1.286259 after 3 attempts
[ INFO] [1558814233.790373357, 39.287000000]: Planner configuration 'arm' will use planner 'geometric::RRTConnect'. Additional configuration parameters will be set when the planner is constructed.
[ INFO] [1558814233.790927439, 39.288000000]: RRTConnect: Starting planning with 1 states already in datastructure
[ INFO] [1558814233.790988817, 39.288000000]: RRTConnect: Starting planning with 1 states already in datastructure
[ INFO] [1558814233.791059312, 39.288000000]: RRTConnect: Starting planning with 1 states already in datastructure
[ INFO] [1558814233.791120209, 39.288000000]: RRTConnect: Starting planning with 1 states already in datastructure
[ INFO] [1558814233.801353892, 39.298000000]: RRTConnect: Created 5 states (2 start + 3 goal)
[ INFO] [1558814233.801434857, 39.298000000]: RRTConnect: Created 5 states (2 start + 3 goal)
[ INFO] [1558814233.801546992, 39.298000000]: RRTConnect: Created 5 states (2 start + 3 goal)
[ INFO] [1558814233.801651412, 39.298000000]: RRTConnect: Created 5 states (2 start ...
It sounds like the joint controllers are not set up correctly, so each joint is freely moving. Can you show us the launch files and controller yaml flies you're using.
Sure thing! Adding them to the OP now.
At a quick glance it looks like your PID controllers need to be tuned. I would expect some of those values to be significantly higher than they are. If you have a look at this page for a slightly different controller it gives you some hints to tune these parameters using dynamic reconfigure. You can also echo the joint effort topics and see what sort of forces the controllers are outputting.
Thank you! Looking into the joint efforts led to me the solution, as found here: https://answers.ros.org/question/1108...