speed up the computeCartesianPath() by using multiple threads?

[boooooo...@gmail.com]

Hi, All:

I want to generate a relatively long cartesian path using computeCartesianPath(), it did its job, but the time it took to compute the trajectory is longer than I expected. Well, one hack around this might be increasing the eef_step so that it doesn't need to compute that many points to cover the whole trajectory, however, I need the execution to be slow so here I deliberately set the eef_step to be small (I tried to set the speed_limit in the joint_limit.yaml but it didn't work and I searched for the solution and it seemed that this is the only hack to make the robot execute the cartesian path slowly).

So here is what I want to do: I want to split the computation in multiple threads, so here is the question, can I call computeCartesianPath() in multiple threads and combine the path manually in the main thread? If not, then how can I speed up the process of computeCartesianPath() when I have a relatively long trajectory?

Many thanks!

Bosch Tang