iCubGenova09 total mass from urdf model does not match the real robot mass #94
Comments
|
@salvi-mattia we should run a second check on all these weights. |
|
CC @S-Dafarra |
|
If you want to check which commit introduced the problem, you can run For a given icub-models commit, you then can check the corresponding icub-model-generator commit in the commit message, see 6d76025 . From that, we can understand what caused the regression. |
|
Very nice feature, indeed! |
2 tasks
|
I wrote this script #!/bin/bash
RED="\e[31m"
ENDCOLOR="\e[0m"
commits=(`git rev-list --ancestry-path cd4b2a6608cea70d0c455b68e18f414cb0820c0c..6d760251473396133daaf75ae7a3b27137da607f`)
for commit in ${commits[@]}
do
echo -e "${RED} ${commit} ${ENDCOLOR}"
git checkout ${commit}
idyntree-model-info -m model.urdf --total-mass #
doneyou should run it inside Here the output |
|
Thanks @GiulioRomualdi for checking. |
|
At this point we have a 3kilos delta between the CAD and RL robot. As me and @DanielePucci said last time at this point we would like to remove this three kilos equally from the 7 big assemblies of the robot [chest, upper legs, lower legs, upper arms, forearms]
|
|
Great, it makes sense to distribute these three kilos around! In fact, we checked more or less the CoM with @GiulioRomualdi and @isorrentino and more or less the CoM was where the model was suggesting. This CoM check, however, could be a further point to be investigated later on, sticking now with the objective of having |
|
A PR alleviating this weight mismatch is expected to land soon. Thanks @salvi-mattia @mfussi66 @ale-git! |
This was referenced Dec 7, 2021
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The iCubGenova09 total mass from the urdf model does not match the real mass of the robot.
Here the one from the urdf model:
and here the reality:
cc @DanielePucci @GiulioRomualdi @paolo-viceconte @S-Dafarra
The text was updated successfully, but these errors were encountered: