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
Inaccuracy in goal poses #458
Comments
Thanks for the in-depth report. From a driver's perspective two things would be important:
Also, your driver launchfile should obviously not print:
If all this is the case, then please check the forward kinematics of your robot using Note: I would not expect any of these to fail. My wildest guess is that the running driver (description) does not use the correct calibration files, as otherwise this error should not exist. |
Apparently, there was a problem with the calibrations files, as the calibration error that you mentioned was present at launch time. On the other hand, I followed the calibrations steps and passed the calibrations files to the launchfile. So I'm going to explore again the calibration procedures. |
Versions
Impact
Some trajectories executed end with a relatively big error from the actual goal.
Issue details
When planning and executing trajectories through MoveIt, for some of the trajectories generated, the reached goal differs up to 2 cm from the actual one.
This situation only happens with the real robots, not in simulation using the same scripts.
Acceptable Goal
Inaccurate Goal
Steps to Reproduce
Generate the calibration files.
Launch the robot driver.
Plan and execute multiple trajectories from the same start state and to the same pose goal.
In order to reproduce this situation I used a simple python script to plan to the same goal multiple times. Many trajectories are generated due to the nature of the OMPL planner. In my case, I get up to 4 different trajectories. For some of this trajectories the (Always happens with the same trajectories) reached goal is far from the actual goal.
Used script:
Final pose results
For the acceptable goal showed in the photos:
For the inaccurate goal:
Trajectory results
For the acceptable goal:
For the inaccurate goal:
Expected Behavior
Getting more accurate goals when planning to the same pose goals.
Actual Behavior
As you can appreciate in the photos the difference between both goals its considerable, but in the final pose recorded with the script there's no such difference.
The text was updated successfully, but these errors were encountered: