WIP: Generated using the latest features from MoveIt

[tylerjw]

Description

This was created by generating using the master branch of MoveIt. There are a handful of files that were not generated (i.e. config/panda_arm_hand.srdf.xacro), I'd like to document how they are created and what their use is in the README.

Lastly, I chose two new poses that are as close to the original ones as I could make and stay out of collision. I didn't include the transport pose because I was not able to make that pose and stay out of collision. I've included images of each pose here:

Lastly, this should be tested with moveit_tutorials to make sure it doesn't break any of those.

Notes to reviewers

Here are the things that need to be done before this is ready:

• Identify and document changes made after generating from MSA
• Discuss new poses
• Test with moveit_tutorials

[v4hn]

The `transport` pose is the configuration used to ship the robot in its original box. That's the main reason why it was part of the configuration. Yes, it's in collision with the current model and this was discussed somewhere before. The MoveIt-compatible trajectory controller can't move the robot to this pose because the internal self-collision check will fail as well. Apparently they have a separate internal routine to move to this pose.

[tylerjw]

I checked through the commits since the last release and this only backs out one change. There is a PR to add the transport pose to one of the files. This specific file I think we can safely remove, this I explain below.

This brings us to a point of discussion. There are two sorts of files in this repo that are not generated by moveit_setup_assistant. The first is xacro files that were copied versions of the old robot description. I believe these can be safely removed.

The next are files that support specific tutorials and were manually added here. An example of that is the lerp_planning_pipeline.launch.xml and related changes. I'm not sure what the right approach is for these sorts of files. I think it makes sense to document what these files are so that in the future if someone needs to generate this package to update it they know what to do about those files.

That or we should move these launch and config files out of this repo and into the tutorials repo to go alongside their respective tutorials. A third option would be to extend the setup assistant to generate these files if they are actually general-purpose launch files that would be useful for other robot setups.

@rhaschke @v4hn do you have thoughts on how we should best maintain these manually added files? I am working through identifying them and trying to validate the functionality of this PR with tutorials.

[tylerjw]

Here is a list of the tutorials on master of moveit_tutorials with my notes. I'll add notes as I work through each one comparing it to the current version of panda_moveit_config.

• Getting Started
• MoveIt Quickstart in RViz
• Move Group C++ Interface
  • The ready state is slightly different causing some of the paths to be changed slightly, but overall this tutorial works. After this is submitted the images and video should be updated.
  • Noticed the generated rviz file doesn't have the rviz_visual_tools gui panel. I'm going to put that file back and note it as not a generated file.
• Move Group Python Interface
  • Python3 migration is needed to run this on Noetic.
  • Tutorial works the same with the new and old configuration.
  • The position of the box using both the new and old configuration is odd.
  • 
• MoveIt Commander Scripting
• Robot Model and Robot State
• Planning Scene
• Planning Scene Monitor
• Planning Scene ROS API
  • Object spawns inside the gripper. See image.
  • The last step "Remove the object from the collision world" doesn't seem to work as the object is not removed. This seemingly broken behavior exists with the old config too.
  • 
• Motion Planning API
• Motion Planning Pipeline
• Creating MoveIt Plugins
• Visualizing Collisions
  • This one is currently broken on the current setup.
• Time Parameterization
• Planning with approximated Constraint Manifolds
• Pick and Place
  • This works as-is but generates a much crazier path because of the new collision model for the robot.
• MoveIt Grasps
• MoveIt Task Constructor
• MoveIt Deep Grasps
• Subframes
  • This is broken with the old configuration with master of MoveIt on Noetic. The object spawns floating in space between the grippers. The grippers should probably be closed to "hold" the cylinder but that is probably not the primary cause of the issues with this tutorial. Each of the steps fails to plan, I'm not sure why. @felixvd have you seen this?
  • 
• MoveItCpp Tutorial
• MoveIt Setup Assistant
• URDF and SRDF
• Low Level Controllers
• Preception Pipeline Tutorial
• IKFast Kinematics Solver
• TRACK-IK Kinematics Solver
• Kinematics Configuration
• Custom Constraint Samplers
• OMPL Planner
• CHOMP Planner
• STOMP Planner
• TrajOpt Planner
• Planning Adapter Tutorials
• Joystick Control Teleoperation
• Realtime Arm Servoing
• Benchmarking
  • Needs https://index.ros.org/p/warehouse_ros_mongo/github-ros-planning-warehouse_ros_mongo which isn't released for Noetic yet. The instructions need to be updated to include building this from source until this is released.
  • I'm confused on how to run the first section called Example. It feels like it is missing instructions for loading the Kitechen1 scene objects. Maybe I did something wrong?
  • Python3 compatibility layer for roslaunch in CMake warehouse_ros_mongo#35 is needed for running on Noetic.
• Integration/Unit Tests

Generated then modified files

• launch/moveit.rviz - restored to state on melodic-devel

Custom Files added for Tutorials

• placeholder

Modifications needed for tutorials

• Update video and images for "Move Group C++ Interface"
• Fix the position of the box in the "Move Group Python Interface" so it is in between the fingers.
• Update MSA tutorial to include instructions for the default positions of the end effector.
• Clarify the "Example" section of "Benchmarking" to include instructions for the collision scene objects.
• Add instructions to "Benchmarking" for building warehouse_ros_mongo from source until it is released to Noetic.
• Add note to tutorials README.md that rosdoc_lite is needed to run the build_locally.sh script.
• Fix "Planning Scene ROS API" tutorial to spawn the object in the gripper and remove it from the world in the last step.
• Fix "Visualizing Collisions"
• Fix "Subframes" tutorial

[v4hn]

The first is xacro files that were copied versions of the old robot description.
I believe these can be safely removed.

This is not the "old robot description", but the srdf file that is split in multiple sub-files. If you drop the ability to load the arm without gripper from the updated package, they should be removed. Some users might be interested in using the arm with other endeffectors, but they should usually generate their own moveit config anyway. So unless the distinction is required for some tutorial, I assent to revert to a single srdf file.

A third option would be to extend the setup assistant to generate these files if they are actually general-purpose launch files that would be useful for other robot setups.

If some files are general-purpose, they should be added to the templates.
The LERP planner is created in the tutorial though, so I would just keep it in the panda config with an appropriate comment at the top of the file.

That or we should move these launch and config files out of this repo and into the tutorials repo to go alongside their respective tutorials.

The planning_pipeline "plugin-like" structure does not allow for this and we discussed some time ago in some issue whether we want to extend it to do so. At least my personal opinion was that it does not make sense to add plugin mechanisms to launch files and people can directly include them in their own launch files.
However, I'm not sure how much overhead that would create for the tutorial's launch file, which should definitely be as simple as possible to support new users.

[tylerjw]

This is not the "old robot description", but the srdf file that is split in multiple sub-files. If you drop the ability to load the arm without gripper from the updated package, they should be removed. Some users might be interested in using the arm with other endeffectors, but they should usually generate their own moveit config anyway. So unless the distinction is required for some tutorial, I assent to revert to a single srdf file.

I believe this functionality is now provided by the description files in the upstream franka package. From what I could tell our versions of these do not differ in functionality from the official ones now. I have yet to find a tutorial that used them, or tested loading without the gripper using the upstream description files.

[tylerjw]

Today I went through and investigates some of the easier to test tutorials. I'll need to work through the rest of them but unfortunatly I found some that don't work with the old config and the master branch of MoveIt on Noetic. Some of these might be easy fixes but I'd don't believe they should block this change and should hopefully be unaffected by this change.

[v4hn]

This is not the "old robot description", but the srdf file that is split in multiple sub-files.

I believe this functionality is now provided by the description files in the upstream franka package. From what I could tell our versions of these do not differ in functionality from the official ones now. I have yet to find a tutorial that used them, or tested loading without the gripper using the upstream description files.

Are we talking about the same files? hand.xacro in this repository provides SRDF specifications, hand.xacro in franka_ros provides URDF specs. These files do not have anything in common aside from their name.

[tylerjw]

Are we talking about the same files? hand.xacro in this repository provides SRDF specifications, hand.xacro in franka_ros provides URDF specs. These files do not have anything in common aside from their name.

I didn't realize this. I figured they provided the same utility. My reasoning was based on searching for files that used these various files in the panda_moveit_config repo and in all cases it seemed like they were referring to the one in franka_description. I'll put those back.

[henningkayser]

Overall, I really think it makes a lot of sense to keep configs up-to-date with the MSA, imo it's just strainge if this is not possible. I'm just not sure what setups we would break by doing this. Of course we should keep modifications like with the srdf files and default poses consistent.

[henningkayser]

Not sure if we should really remove this

[JafarAbdi]

Note that I disabled the collision between panda_link8 & panda_link6 + panda_link7 otherwise hand movement is very restricted causing some of the tutorials to not work

[rickstaa]

I created frankaemika/franka_ros#162 to make the Franka Emika team aware of the restrictive hand movement problem. @JafarAbdi please free to add any points I missed or misunderstood in your original explanation to the issue.

[rhaschke]

Thanks a lot, @rickstaa, for pushing this forward. A new release of panda_moveit_config is long overdue!
I had a look at your branch rickstaa#26 and cleaned it a little bit, following these rules (see here for the result):

• commit messages shouldn't include PR IDs from your fork
• fixups due to changes in MSA behavior were squashed directly into the base commit

Regarding your questions:

• I suggest including examples for other planning pipelines even though these pipelines are currently not operational or released. Otherwise, it will become hard to re-add those files later from scratch.
• For the Gazebo simulation, I suggest to use franka_gazebo if that's closer to the real robot. Does that require changes in the tutorials as well?
• If the existing files differ to the MSA-generated ones: Keep the generated ones, if generic files (e.g. planning pipeline settings) are concerned. But keep robot-specific settings like joint limits etc. For those, MSA can only use the defaults from the URDF.
• @Le-Li-17 please have a look at the long list of PRs filed to Franka. We strongly believe that the URDF model released with version 0.8.1 still has severe issues strongly limiting the motion range of the simulated robot. See Update SRDF for collision model #35 (review)

[rickstaa]

@rhaschke Thanks a lot for helping me implement this new release.

Thanks for rebasing my commits they look a lot cleaner! I will use https://github.com/rhaschke/panda_moveit_config/tree/noetic-devel-rickstaa and adhere to these rules moving forward. Are these rules written down somewhere in a contributing.md. Or are they requested when people make a pull request?

I suggest including examples for other planning pipelines even though these pipelines are currently not operational or released. Otherwise, it will become hard to re-add those files later from scratch.

Are you okay with adding them commented out since otherwise, they throw errors (see https://github.com/rickstaa/panda_moveit_config/blob/f4e1c046d7ca88d9e06f0969ed81cfda62fb2d0b/launch/move_group.launch#L64-L70).

For the Gazebo simulation, I suggest to use franka_gazebo if that's closer to the real robot. Does that require changes in the tutorials as well?

Good point I did not yet look into this. I check and track all the changes that need to be performed to the moveit_tutorials in rickstaa#26.

@Le-Li-17 please have a look at the long list of PRs filed to Franka. We strongly believe that the URDF model released with version 0.8.1 still has severe issues strongly limiting the motion range of the simulated robot. See Update SRDF for collision model #35 (review)

I'm trying to fix this with @frankaemika. frankaemika/franka_ros#162 is related to this pull request might people want to chip in.

[rhaschke]

I will adhere to these rules moving forward. Are these rules written down somewhere in a contributing.md?

I just formulated these rules in a common-sense fashion to answer you questions in a generic way. I hope other maintainers agree.

I suggest including examples for other planning pipelines even though these pipelines are currently not operational.

Are you okay with adding them commented out since otherwise, they throw errors.

I wouldn't include them by default into move_group.launch, but consider them as optional custom pipelines to be included manually via the pipeline argument.

[rhaschke]

Sure! Even though the planner is failing, the config is probably still fine.

[rickstaa]

@rhaschke I think I just created the last pull request needed to release the Noetic branch (i.e. rhaschke#8). Just to be sure, tomorrow I will perform some tests on the real robot to see if rickstaa#54 is still needed in ROS Noetic. After that, the Noetic branch is ready to be released when the following upstream PR are merged:

• Moves finger collision geometries to the finger joints frankaemika/franka_ros#151 needs to be merged to get rid of the Link panda_rightfinger has visual geometry but no collision geometry. Collision geometry will be left empty. Fix your URDF file by explicitly specifying collision geometry. warning.
• Adds position and velocity motion generators to FrankaHWSim frankaemika/franka_ros#181 needs to be merged for us to use the position_joint_trajectory_controller.
• Fixes franka_gripper_sim GripperCommand action frankaemika/franka_ros#173 needs to be merged before MoveIt can correctly control the gripper. Currently, MoveIt can only open and close the gripper but not set a desired width.
• Adds 'use_collision_mesh' arg to the robot description frankaemika/franka_ros#154 could improve the self-collision behaviour. When this pull request is merged, we can 'use_collision_mesh' arg to let MoveIt plan with the more accurate collision meshes. In that case aa24fae is then not needed anymore. If we however decide to do this we have to keep in mind that MoveIt can plan paths that are not possible on the real robot.
• The Noetic branch has been created using the moveit_setup_assistant found in tree/82378877879f7d40727819b1c3345eeb4308d98b. The current MoveIt debian package does not include the pilz_industrial_motion_planner. As a result, MoveIt first has to create a new release before it can be used with the debian packages.

Migration guide

For future reference, the following steps were performed to release the Noetic branch:

1. I created the initial panda robot configuration using the moveit_setup_assistant (see 8566b70). In this branch, tree/82378877879f7d40727819b1c3345eeb4308d98b was used. Similar to WIP: Generated using the latest features from MoveIt #74 I changed some posed to prevent self collisions.

2. I modified the panda.srdf file that is created by the moveit_setup_assistant such that we can programmatically enable the gripper (see f0c2bca). This needs to be done, since the moveit_setup_assistant currently doesn't propagate xacro arguments you supply to the urdf.xacro file (see this issue). We therefore need to manually add a way to enable or disable the gripper. This can be done by creating a panda.srdf.xacro file.

3. I replaced the sensor_3d.yaml with the example files in the perception_pipeline tutorial. Also edit the sensor_manager.launch.xml file to point to the right configuration files (see 2aaa01d).

4. I added the rviz_tutorial flag. While doing this, I made sure that I copied the moveit.rviz and empty.rviz files from the previous ROS version. Additionally, I added the moveit_visual_tools package as a dependency (see 1d55c98).

5. Similar to df0cbd8 I had to disable the collision checking between joint{5, 6} and joint8 since the new collision geometries that were added to https://github.com/frankaemika/franka_ros/releases/tag/0.8.0 are to coarse (see aa24fae). This should be fixed upstream by, for example, by improving the collision geometries or by giving users the ability to use meshes instead (Adds 'use_collision_mesh' arg to the robot description frankaemika/franka_ros#154).

6. I added the panda_control_moveit_rviz.launch which can be used to control the real robot (see 36d7c9f).

7. I added the TrajOpt files that were found in the melodic-devel branch to the repository ( see bc2bbc7 and 88fb278).

8. I added the STOMP files that were found in the melodic-devel branch to the repository ( see 133d524) .

9. I updated the join_limits.yaml file that was created by the MSA to enforce jerk limits. This was done by adding acceleration limits, since MoveIt does not yet support jerk limits (see 89d9b44).

10. I added the files of the LERP planner that is created in the creating-moveit-plugins tutorial (see 85756cd). These files were copied from the melodic-devel branch.

11. I added the demo_chomp.launch and ompl-chomp_planning_pipeline.launch.xml files needed by the chomp_planner_tutorial tutorial. I also updated the chomp config file to the one found in the melodic-devel branch (see 4c390ac and c2ccca2).

12. I added the demo_gazebo.launch launch file which spawns a simulated Panda robot in Gazebo (see c46c81c). This launch file uses the new franka_gazebo package that was introduced in franka_ros v0.8.1.

Files that were manually created

Config folder

• panda.srdf.xacro
• hand.xacro
• panda_arm.xacro
• lerp_planning.yaml
• stomp_planning.yaml
• trajopt_planning.yaml
• panda_controllers.yaml
• panda_gripper_controllers.yaml

launch

• demo_stomp.launch (Was not present in the melodic-devel branch).
• demo_chomp.launch (Not used in the moveit_tutorials).
• demo_lerp.launch (Used in the moveit_tutorials).
• lerp_planning_pipeline.launch.xml
• ompl-chomp_planning_pipeline.launch.xml
• panda_gripper_moveit_controller_manager.launch.xml
• run_benchmark_trajopt.launch
• stomp_planning_pipeline.launch.xml
• trajopt_planning_pipeline.launch.xml

Moveit_tutorial changes

I found the following changes that need to be performed to the moveit_tutorials:

• setup_assistant_tutorial.html:
  • Should have updated pictures (i.e., menu order was changed).
  • Panda description file was changed to be /franka_description/robots/panda.urdf.xacro the hand is now enabled by using the hand:=true xacro parameter.
  • I think the documentation should mention the panda_finger_joint2 as the passive joint.
  • The virtual joint doesn't appear to be needed in the panda_arm group. Maybe remove it as it will cause errors when using a gazebo simulation?
• Maybe the [panda_control_moveit_rviz.launch] file should be documented in the documentation.
• The demo_chomp.launch file was added to the panda_moveit_config repository, but this file is not used in the moveit_tutorials.
• The lerp planner documentation talks about a lerp_example.launch file but this file is not included I think this should be changed to refer the demo_lerp.launch file.

Moveit_setup_assistant changes

I found several improvements that can be made to the MSA.

• Maybe we can add the ability to use action services as controllers (i.e. https://github.com/rickstaa/panda_moveit_config/blob/47073b308a66a322146c7544f943533575f99d9e/config/panda_gripper_moveit_controllers.yaml#L14-L20).
• Maybe we should move controllers_list from ros_controllers.yaml into its own file as these are only needed for MoveIt while the settings in the ros_controllers.yaml file should only contain settings related to the ROS control (see Adds the panda_control_moveit_rviz launch file rickstaa/panda_moveit_config#45 for a example).
• The code that loads the robot_description in the gazebo.launch` template from:

  <param name="robot_description" textfile="$(arg urdf_path)" />

  to

  <param name="robot_description" command="$(find xacro)/xacro $(arg urdf_path) hand:=$(arg load_gripper) gazebo:=true"/>

  if a xacro file is supplied by the user.
• As the https://github.com/frankaemika/franka_ros was changed, the code that creates the gazebo simulation should be adjusted. The https://github.com/frankaemika/franka_ros repo now includes all the components to create a gazebo simulation, but loads the robot onto the /panda namespace. As a result, the current templates don't work anymore. We should apply the changes found in Fixes the gazebo simulation rickstaa/panda_moveit_config#40 to the templates in the MSA.
• The STOMP templates should be added to the MSA.
• The stomp_demo and chomp_demo files are not created by the MSA. Maybe we should add these files?
• The ompl-chomp_planning_pipeline.launch.xml should be added to the MSA (see https://ros-planning.github.io/moveit_tutorials/doc/chomp_planner/chomp_planner_tutorial.html#using-chomp-as-a-post-processor-for-ompl).
• The CHOMP algorithm contains several commented out parameters that might be removed (see https://github.com/ros-planning/moveit/blob/d57721cf8503fe893347196cc1e1f9249ec57777/moveit_planners/chomp/chomp_motion_planner/include/chomp_motion_planner/chomp_parameters.h#L81-L88).
• Maybe we also want to add the remaping found in Remaps move_group 'joint_states' to 'joint_states_desired' rickstaa/panda_moveit_config#54 into the moveit_setup_assistant template.
• The CHOMP ridge_factor parameter should be changed to 0.0 since this is the default.
• The CHOMP enable_failure_recovery parameter should be changed to false since this is the default.

Other upstream problems that still exist

• I did not test the workings of the TrajOpt planner since it did not build on my system (see Trajopt with no dependency to tesseract moveit#1626 (comment)).
• The STOMP planner seems to fail on ROS Noetic due to a upstream bug.

Other things we need to do

• We need to re-test all the moveit_tutorials with the new Noetic branch.

[rhaschke]

Big thanks for your work, @rickstaa 🎉
I have pushed the branch to https://github.com/ros-planning/panda_moveit_config/tree/noetic-devel

Could you please open a new issue (e.g. titled Noetic release) with all the info in #74 (comment)?
Closing here.

[simonschmeisser]

Note that STOMP moved to a separate repo some while ago and the issue was just fixed: ros-industrial/stomp_ros#26

[rickstaa]

@rhaschke I created #89. Feel free to request changes to the TODOS that are in there.

[rickstaa]

@simonschmeisser Thanks a lot for letting me know I will re-test the STOMP planner.

[rickstaa]

@simonschmeisser Your right the upstream issue has been solved thanks for letting me know.