iCubGenova11 S/N:050 – The left arm oscillates in certain configurations, probably due to the shoulder roll

[xEnVrE]

Robot Name 🤖

iCubGenova11 S/N:050

Request/Failure description

While testing a demo today I noticed that in certain configurations the left arm oscillates a lot making the robot possibly unsafe. At a certain point the should roll went in HF due to i2t.

FYI cc @enyen @GiuliaP

Detailed context

I was running a demonstration where the arm is controlled in Cartesian space. Along the planned trajectory there are configurations in which the arm oscillates a lot. The demonstration is not new and has been executed lots of times, hence we know it is safe.

To me the problems seems very similar to what we had in https://github.com/robotology/icub-tech-support/issues/1488#issuecomment-1406316146. Hence, probably the encoder is not well connected.

Additional context

I am attaching the yarprobotinterface log.

log_icub-head_yarprobotinterface_2668.txt

How does it affect you?

@gabrielecaddeo would need to use the left arm for a demonstration this Friday. Without the arm we can partially execute the demonstration. Thank you

[GiacomoVassallo13]

Hi @xEnVrE . The left shoulder roll locked motor has been restored. The malfunction was caused to the loose srews of wave generator of harmonic drive. The screws has been thight and the harmonic was cleaned and regreased. The joint it' s ready to test. I attend the feedbacks

cc: @Fabrizio69 @sgiraz

[xEnVrE]

Thanks @GiacomoVassallo13

At the moment we cannot run the robot due to the following errors:

Attaching the full yarprobotinerface log just in case:

log_icub-head_yarprobotinterface_2141.txt

cc @sgiraz @AntonioConsilvio

[xEnVrE]

I tried again. It seems that the system detects the fault button as pressed - as if the fault line was interrupted somewhere:

Full log: log_icub-head_yarprobotinterface_2521.txt

[sgiraz]

Hi @xEnVrE,

@AntonioAzocar is going to check for faulty cables... stay tuned!

[xEnVrE]

We have tested the motion of the shoulder roll and it was fine. Closing.

Aside: the robot did not started because of a malfunctioning torso IMU. I am going to open a new issue specifically for that. For the moment being we are commenting the relevant part inside the icub_all_no_legs.xml file inside $ROBOT_CODE/src/robotology_superbuild/src/robots-configuration/iCubGenova11.