-
Notifications
You must be signed in to change notification settings - Fork 333
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add wrench offset for admittance controller #1249
Add wrench offset for admittance controller #1249
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In general, the changes look good to me. I've left some minor comments. Please take a look
Co-authored-by: Sai Kishor Kothakota <[email protected]>
@firesurfer can you remove the draft status on the title? |
@saikishor I removed the draft status. |
Any of the other reviewers willing to take a look at this ? |
Do you have any ideas for a test for this by any chance? |
Co-authored-by: Sai Kishor Kothakota <[email protected]> Co-authored-by: Bence Magyar <[email protected]>
The way I tested it on hardware was: Load the controller and then command 1N into a single direction. One will then see a motion of the arm in that direction. I guess it would be possible to implement something similar in an unit test. |
@destogl When browsing through the tests I found this:
Apparently there was some support for a force command planned but not implemented? Is this right? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Generally, great!
What I would like to know is, do we want to enable this input even admittance controller is in a chain? As you did right now. I am more on a yes
side, but would like to hear opinions.
{ | ||
if (msg.header.frame_id != admittance_->parameters_.ft_sensor.frame.id && !msg.header.frame_id.empty()) | ||
{ | ||
RCLCPP_ERROR_STREAM( |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Why don't we transform the wrench if it is in another frame to the sensor frame?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I had some very weird results in the past when transforming wrenches with TF (e.g. non zero torques afterwards when only forces where given). Additionally I don't think we have a tf buffer/listener in the controller at the moment and why add this additional overhead?
…:firesurfer/ros2_controllers into feature/admittance_wrench_target_rolling
So I just updated the code, that it compiles again in jazzy and rolling. EDIT: Apparently it fails in the CI. I tested it locally on my system with an up to date Jazzy installation and it compiled just fine. |
This reverts commit 06ff415.
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## master #1249 +/- ##
==========================================
- Coverage 83.64% 83.60% -0.04%
==========================================
Files 122 122
Lines 10985 11006 +21
Branches 932 933 +1
==========================================
+ Hits 9188 9202 +14
- Misses 1488 1494 +6
- Partials 309 310 +1
Flags with carried forward coverage won't be shown. Click here to find out more.
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I never used this controller so I don't comment on the changes itself but it seems to be valuable acc. to previous reviews.
But please:
- Update the controller's documentation
- Add this new feature to the release notes.
So the only outstanding comment is the one from @destogl about the force command transformation. I don't think this is a blocker at the moment as it could be changed rather easy afterwards. |
a6c1c0c
into
ros-controls:master
This PR adds an additional WrenchStamped subscriber to the AdmittanceController in order to allow specifying a Wrench offset.
The basic idea behind this PR is to simply add the offset to the measured force values which has the controller to produce an offset into the desired direction. I tested it on an UR16e where it seems to work quite well.
Is this a desired feature? What do you think of it? Perhaps there are better ways to implement this feature in this controller.