Clarification on the zero configuration of *_index_add of the simulated Mk5.1 hand

[xEnVrE]

I am opening this to ask a clarification about the behavior of the *_index_add joint of the Mk5 .1 hand.

According to this comment, it was decided that the zero position of the index adduction is the one where the index is more distant from the other fingers, i.e.:

It seems that such new behavior has not been transferred correctly in the simulation:

index_add = 0

index_add = 15

I guess we should:

• invert the direction of the axis of rotation
• change the default configuration at startup in Gazebo

cc @Nicogene @pattacini @fiorisi

[pattacini]

Good catch @xEnVrE 👍🏻

@Nicogene @mfussi66 please read up on the issue from https://github.com/icub-tech-iit/ergocub-design-lowerarm/issues/157#issuecomment-1660679203 onward for more info on that.

You were off. I tried to find out why we ended up with a different convention wrt the iCub hand but unsuccessfully. At this point, we ought to conform to the latest standard.

[Nicogene]

Since creo2urdf assumes that in the simulation model, all the joints are in their zeros, I need that @Mick3Lozzo or @Lawproto to change the rest position of the index_add joints in the simulation model.

Or actually, maybe it is the chance for changing the range to -3 to 12 degrees, see #141

What about the thumb_add ? Also that has to be changed?

[xEnVrE]

Or actually, maybe it is the chance for changing the range to -3 to 12 degrees

I am a bit confused, isn't it an issue related to Mk5.0?

What about the thumb_add ? Also that has to be changed?

I would say no, but others might comment on this. The behavior of the thumb adduction is at the moment similar to that on the old iCub hands, i.e. 0.0 -> thumb within the plane of the palm, max -> thumb ~orthogonal to the palm or the like

[pattacini]

Since creo2urdf assumes that in the simulation model, all the joints are in their zeros, I need that @Mick3Lozzo or @Lawproto to change the rest position of the index_add joints in the simulation model.

Let's do it.

[Lawproto]

@Nicogene @pattacini to sum it up, I inderstood I have to update the ergoCub 1.1 sim model with the new index position.

It this is the case, the index_add joint rom spans from 0 (figure) to 15°.
cc @fiorisi

[pattacini]

Yep 👍🏻

[Lawproto]

@Nicogene as discussed via call, I committed the sim_ergocub_1-1.asm in this position.

Now it's ready to fire some guns.

cc @pattacini

[Nicogene]

After @Lawproto fix, here is the PR that port those changes in the urdf:

• ergoCubV1_1: correct index_add joint #173

0 deg

15 deg

cc @fiorisi @pattacini @xEnVrE