This week, I controlled a robot with my brain. I’m gonna say it again, so it can sink (and cause I like saying it).
I controlled a robot with my brain.
This final project for my Neural Engineering lab. A lego robot, controlled by neural signals (we’re using EMG), that has to be able to navigate a maze. The EMG circuitry is connected to the computer and our software (labview, expect a separate post on that at some point) checks for spikes indicating neural signals being sent to the muscle. The amperage of the computer’s output is too weak to actually drive the robot’s motors, so we use the signal to flip a transistor that’s hooked up to 9 volt batteries that actually drive the motors. This happens every time the system sees a spike.
On Wednesday we had our first live test (using real EMG signals instead of a recording). We set it up to drive both motors at the same time (we haven’t worked out driving the wheels separately yet). Every time I flexed my muscle, the robot would scoot forward. Well, not quite. Our system is nowhere responsive enough to feel that transparent. You had to really work to get the spike high enough that the motors would run, and sometimes they’d keep running for a while after you stopped flexing. But, man. It. Was. Awesome.