Cool project
one of the things we had the option of trying in our biorobotics course was brainwave control. Only thing was, that it worked best when blindfolded and in a quiet room!! Obviously not reliable.
And you are right, FIRST would have to change there rules. There is not enough processing power in the IFI controller for me to use neural networks and fourier processing like I had planned. The original idea was to run it through a computer to do all the processing, but I decided on a simpler solution, but it isnt competition worthy... yet