The GA was based on an open loop simulation. I tuned a digital canvas / painting simulation by hand and then evolved the paintings in simulation. After evolvi

The GA was based on an open loop simulation. I tuned a digital canvas / painting simulation by hand and then evolved the paintings in simulation. After evolving the painting I would execute the painting with a robotic arm. I did some preliminary work on closing the loop: using the actual output of the painting to tune the simulation but didn't get too far. A paper on my methods / simulation and results: http://carlos.ag/GA2008paper.pdf http://carlos.ag/GA2008paper.pdf