You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Similarly to #473, I am trying to understand the code of the coupling handlers for the finger adduction/abduction.
From physical joints space to controlled DoFs space
decouplePos should move from the physical joint space (i.e. the four joints associated to index, middle, ring and little) to the controlled DoFs space (i.e. the iCub {l, r}_hand_finger, i.e. joint no. 7).
Then, shouldn't it be something like this? This because the range spanned by the unique DoF {l,r}_hand_finger is three times the range spanned by the link associated to the little finger (or, alternatively, that of the index or that of the little).
From controlled DoFs space to physical joints space
decoupleRefPos should move from the controlled DoFs space (i.e. the unique DoF {l,r}_hand_finger) to the physical joint space (the four joints associated to index, middle, ring and little)
Then, shouldn't it be something like this? This because the best we can do is to associate one third of the range spanned by {l,r}_hand_finger to each joint (out[m_coupledJoints[1]] is set to the 0.0 as the link associated to the middle finger does not move on the real robot).
The situation is not really clear to me here. Probably as a start it would be useful to have a clear definition/agreement with names of the physical and controlled joints, see robotology/icub-models#30 . cc @kouroshD
Similarly to #473, I am trying to understand the code of the coupling handlers for the finger adduction/abduction.
From physical joints space to controlled DoFs space
decouplePos
should move from the physical joint space (i.e. the four joints associated to index, middle, ring and little) to the controlled DoFs space (i.e. the iCub{l, r}_hand_finger
, i.e. joint no. 7).gazebo-yarp-plugins/plugins/controlboard/src/ControlBoardDriverCoupling.cpp
Lines 427 to 432 in 8c056e2
Then, shouldn't it be something like this? This because the range spanned by the unique DoF
{l,r}_hand_finger
is three times the range spanned by the link associated to the little finger (or, alternatively, that of the index or that of the little).From controlled DoFs space to physical joints space
decoupleRefPos
should move from the controlled DoFs space (i.e. the unique DoF{l,r}_hand_finger
) to the physical joint space (the four joints associated to index, middle, ring and little)gazebo-yarp-plugins/plugins/controlboard/src/ControlBoardDriverCoupling.cpp
Lines 455 to 464 in 8c056e2
Then, shouldn't it be something like this? This because the best we can do is to associate one third of the range spanned by
{l,r}_hand_finger
to each joint (out[m_coupledJoints[1]]
is set to the0.0
as the link associated to the middle finger does not move on the real robot).The text was updated successfully, but these errors were encountered: