The shiffman videos reminded me of the kinect sitting idle in the center of my InMoov.
I have tried to follow his examples on my PC and had some troubles making the kinect work again (driver issues) and make it run on my PC (his examples are shown on a mac) and finally succeeded with it.
Now I wonder wether I could run the power hungry image and depht analysis on my PC with Processing and send specific move commands to my bot where I run MRL on a ubuntu/odroid device.
Or does this scenario not make sense and it would be better to have a fast windows laptop as the MRL machine? And how could I combine Processing with MRL?