5X5 Episode 22

Researchers allow monkey to control robotic arm with its mind.
S: This is the SGU 5x5 and the topic for today is: monkeys learn to control a robot arm with just the power of their mind.

R: Birds stare into a mirror and chirp.

(laughter)

S: This was research done at Duke University, project leader is Miguel Nicolelis and what they did was they attached a interface with monkey brains, they put the monkeys in a box so they couldn't move and attached to them was a robotic arm and the monkeys very quickly learned to control the robotic arm so that it could reach out, grab a piece of food and then deliver the food to their mouths. So they were essentially able to feed themselves without moving physically, but just mentally using the robotic arm.

B: This has obvious potential applications for prosthetics, for people that have lost an arm or a limb. Remote telerobotics or remote control of robot devices. Steve, earlier you mentioned that--when we were discussing this--potentially the possibility of accessing information using this kind of type of interface has some interesting ramifications.

S: Yeah, this is a tremendous proof of concept but what this demonstrates is the ability for information to go from the brain to a computer board or a computer chip - to a computer interface. And this is hooked up to the motor cortex of these monkeys. So it's the part of the brain that's used to moving something physical - moving an arm, for example.

B: Yeah, this a key distinction because other prosthetics that I've seen, they would actually attach the prosthetic to say, your pectoral muscle. You would flex your chest muscle of, say, the side of your body without an arm and that muscle would then - you would use it to control the prosthesis-

S: That's right.

B: -and, so this is the first- Is this the first one Steve, that was - that is in fact connected to the motor cortex and not to muscle?

S: That's correct. Yeah, I'm pretty sure that this is the first one, although there were previous experiments where - actually, I mentioned the researches at Duke University, they actually did the prior research a few years ago where monkeys were able to control a cursor on a screen. This new research just published in "Nature" online was done at the University of Pittsburg by Dr. Andrew Schwartz and they were the first ones to connect the motor output--the connection to the brain--to an actual external device, a robotic arm. So this was just taking this one more step in their proof of concept. You're right. I mean the obvious extension of this is the bionic man type of robotic prosthetics. Giving legs to somebody who is paralyzed from the waist down, for example. Obviously they wouldn't have to replace missing limbs, it could be an exoskeleton over their legs that is controlled with their mind.

E: Another neat practical application for something like this is to have a robot go into a dangerous situation, disarm a bomb, go into combat - something like that and you have somebody at the other end just with a wire hooked up to their mind being able to control everything that goes on.

S: Yeah, the question would be, could this interface get so sophisticated that it would provide superior control to just a joystick or manual interface, like we have now.

B: Also, there's another application I've been thinking of. Guys, tell me what you think of this. Kind of extending the whole prosthesis idea is increasing human bandwidth in a sense, so that not only could your hands control the device but in conjunction with your hands or even, say your legs, say when you're flying a helicopter. But your mind - you throw the mind into the loop at the same time, which could kind of add another dimension to the task. Say it's already maxed out because you only have two arms and two legs and you could add another dimension to the task so that it can become even more - you could perform even more complicated, say, maneuvers than you would normally-

S: Hmm.

B: -because you just don't have enough limbs to do all these controls.

M: So you can have a Dr. Octopus thing going on.

B: There you go, hey-

R: That would be rad.

B: I like it.

R: Even I admit that.

S: Of course, the next place to go with this type of research is to get the information to go the other way, from the computer to the brain. This could be useful for providing sensory feedback. So imagine controlling this disembodied robotic arm that you can't feel. That'd be kind of weird. But imagine if you felt it as if it were-

E: Hmm.

S: -your own arm. It would giving you sensory feedback to make it seem like you actually possessed it, you owned it. That - then the amount of control over that limb could be truly natural and genuine, like - as much as you control your own limbs.