Oky guys I uploaded all the material here.
My Phd is about social robotic so I will keep it updated pretty much every week.
Cheers. 
Hi, Epokh.
That’s very cool, thank you for sharing your project with us. Are the robots learning as they drive around, so that their behavior changes over time? If so, I think it might be fun to see them drive around in a fully enclosed arena over a longer period of time (maybe with the video sped up) so that one can more readily see the behavior change. It also might be interesting to see them on a surface that has multiple food patches. Are they rewarded for finding the center circle?
- Ben
Yes I’m doing longer video to show the difference.
And yes I’m printing another playground with multiple patches.
There’s no internal reward signal to get food, but basically they learn by use:
hebbian synapses they become stronger when the the grey area is followed by black area.
I will test other algorithms as well such as self organising maps to learn some motion behaviour.