Nao "real-time" control and inverse kinematics

asked 2020-04-16 12:06:17 -0500

triguz gravatar image

updated 2020-04-16 12:07:25 -0500

Hello,

I'm a student and I'm doing a university project in which my goal is to control a nao robot using a Oculus Rift. I don't have any experience with ros but I managed to set up a ros master plus various ros - NaoQi driver nodes in a linux machine and ROS# on unity to allow comunications between the two. Now if I understood correctly I should be able to take the position of my joysticks in unity, convert this position to a ros-readable position, send it to ros and with the help of a inverse kinematics library, generate a set of trajectories that will be sent to the robot to execute.

Specs:
- Ubuntu 14.04
- Ros Indigo

If my reasoning is correct I have a couple of questions:

1) What type of data should be the position that i sent out from unity?
2) How can i tell the robot to reach a defined point?
3) Should I use Topics to comunicate with ros? Or is it better to define actionlib or services?
4) What library/solver can I use to calculate inverse kinematics? The aldebaran cartesian control as far a I know is not reliable enough, I saw that moveit ikfast can be used but i'm not sure how or if it's even possible to use it in my case (without moveit), and in case how would i comunicate with it?

I'm sorry if my english isn't perfect, I'm not a native speaker, if something isn't clear please let me know,
All help is appreciated.

edit retag flag offensive close merge delete