Creating a Constraint after a Collision - orientations unclear

Official Python bindings with a focus on reinforcement learning and robotics.
Post Reply
Posts: 1
Joined: Mon Aug 03, 2020 12:38 pm

Creating a Constraint after a Collision - orientations unclear

Post by Passai »

Suppose I have two robots A & B, consisting of the links (a1,a2,..) & (b1,b2,..). When A touches B's link "bi" with its link "aj", I would like to create a point2point-constraint, when they separate, to keep them glued together.
Unfortunately, I cannot really follow the instruction of the quickstart guide as they are either vague or I am missing something.
How can I get the following values? Which orientations do I have to querry with getLinkState and what transforms do I have to use?

a) jointAxis: Does it matter for a point2point-constraint?
b) parentFramePosition
c) childFramePosition
d) parentFrameOrientation
e) childFrameOrientation

Additional Information:
The robots are usually initialised with an orientation different from RPY=(0,0,0). The sticky links of A are spherical, the links of B that A shall be able to stick to come in all primitive shapes.

Any help, suggestions and ansatz are greatly appreciated. Thank you.
Post Reply