Revision history [back]

I agree the TF system is a very good way of solving this challenge. When I've worked on similar things in the past we defined a static transform from the objects body frame to the end effector grasp location. Then keep a record of the location of the object in the world frame and publish the TF from the world to the objects body frame.

The combination of these two transforms means you can do a TF lookup of the grasp pose of the object in the world frame at any time, as well as having a TF of the objects pose as well.

Hope this makes sense.