Monkey's Thoughts Make Robot Walk

geekbits writes "For all those who have at one time or another been too lazy to get up off the couch and go to the fridge and get a beer, heat up some pizza, or change the channel when the remote is missing, we may be one step closer to being able to keep our tushes parked just a little while longer. There may also be some slightly more noble implications here. According to an article in The New York Times, in an experiment at Duke University, a 12-pound, 32-inch monkey made a 200-pound, 5-foot humanoid robot walk on a treadmill using only her brain activity. She was in North Carolina, and the robot was in Japan."

Re:monkey business

[Reivec]

I think the distance is easy to explain, and I doubt it was for show. The research seems to originate from Duke. They likely decided they wanted to see if a monkey could control a humanoid robot but making a robot to see so was outside of the scope of what they were trying to do, they are just making the interface. So they searched out some other team making humanoid robots (which Japan seems to have a lot of). It was likely much easier to setup an internet link to connect the two groups as opposed to meeting in the same location. To meet you would have to move a lot of people and a lot of equipment, all of which would be a customs nightmare. Moving monkeys back and forth over international borders probably requires a lot of checks and paperwork and what not, just as I am sure moving research technology does as well, especially something that large.

To sum up, it was a hell of a lot cheaper and faster that way.

Re:yet another...

[moosesocks]

Sadly, real news more and more seems to resemble The Onion these days.

Perhaps even more alarmingly, quite a few of their more outlandish stories have actually come true several years later.

(This [theonion.com] being one of the funniest such stories...)

Combination of previous works

[DrYak]

This is a combination of previous works.

Monkey mind reading has been done before.
Monkey controlling a robotic arm has been done before too, and as far as I remember, the monkey even got it to the point of controlling the robotic arm without moving herself.
Remote controlling of robots has been done before (trans-atlantic surgery operation, the surgeon operating the robot in the US and the patient being in Europe).
And as pointed by other /.ers, research on walking robot seems so common in Japan that it's probably the second national sport (right after "girls in school uniform fetishism" ^_^ )

What's the point of this study ? Combining all this together.

- It's the first time brain waves are used to control a movement as complicated as walking.
Thus, this is a technical demo that brings closer hopes for paralysed patients. (As a different solution than spinal nerves regrowth).
--> (The previous experiments where robotic-arm only, thus potential application only to amputees).

- It's the first time that brain-waves remote controlling is attempted.
The inputs are much more complex and much more abundant compared to the current joystick-controlled robots.
This technical demo proves that the latency and bandwidth can cope with brainwave-control, although with a lag that maybe won't be short enough for reflex based movements.
But it is still opening interesting possibilities :
Just replace the monkey with a scientist and the treadmill environment with either some dangerous environment or some miniaturised one or a remote place where the scientist expertise is not readily available.
--> Current controls of robots (like the one used in surgery) are joystick based. Although there are still a lot of movement that can easily be performed with such controls, there are some limitation. Natural movement that are mapped to a robot through brainwave control could bring much more agility.
Also a lot of additional things have to be controlled in a surgical robot (camera motions for example). For now they are still controller with the same pair of joystick (because, you see, a surgeon has only 2 hands to hold them) and using a combination of pedals to switch what the inputs are controlling.
With such brainwave-control technology, other movement could be mapped to the camera control (I think head motions could be the most natural ones) leaving the hand motion free to continue controlling the instruments.

So, no there are no revolutionary new technologies involved here.
But its a new combination of technologies that represent a nice step toward very promising applications.

It's using CPGs, just like the monkey's real legs.

[Anonymous Coward]

The signals in the brain are calling to learned motion patterns stored remotely in the ganglia (little brain-like things attached to the spine). A good computer analogy would be shader programs--the software makes very vague system calls (render these triangles), which cause separate (shader) programs to actually draw the graphics from inside the video card.

What these guys are doing is capturing the "system calls" from the brain, and sending them to *their own* central pattern generators in the robot. According to TFA, these CPGs were trained by data-mining a correlation between the monkey's motion (picked up with standard motion-capture methods) and the firing of its neurons. So, the first step is actually to train the robot to obey the monkey's brain.

What happens next is the really interesting part, though. The monkey's brain quickly realizes that it is controlling two separate things (the monkey and the robot), and moves the control for the legs to different neurons so it can control them separately. This is why, when they stopped the monkey's treadmill, it was able to keep controlling the robot without moving its own legs.

This research isn't all that new, though. Something similar was done a while back (maybe a year ago?) where they had a monkey controlling an extra arm through the same type of system. Before that, there was a study on epilepsy patients that involved the same brain sensors, where humans were able to control a mouse cursor after trying for about one minute. IMHO, this problem was completely solved long before this project; electronic control outputs from the brain just happened to be one of those things that are way easier than they look.

What someone really needs to do now is get an electronic brain *input* that works this well, and doesn't cause seizures or degradation over time. Once that's done, then some really amazing stuff will start to be possible.

Besides just being able to control machines with your thoughts, or having an in-head network connection, or some other such nonsense, this is one of the technologies that could truly make it possible to live forever, for people who are alive today. If one's whole cerebral cortex is wired to a computer that's emulating (get this) more brain, then it's going to do exactly what it does with the robot, and start using it. Since most of the brain naturally tries to maximize its redundancy, you'd get a hybrid digital-biological brain that could be trained over time to work both together and separately. Assuming that they're together when you "die", then that one thread of consciousness will simply continue to exist on the digital side, and you will survived death. At this point, you can get in your robot and go do stuff, or run around in some kind of digital world. Of course, computer hardware today would not be able to do this, but in 10-15 years it could very well be possible using just consumer stuff (and will definitely be possible using a server).

As a side bonus, it will also cause a very entertaining political flamewar, when the fundies realize that there's a machine that traps your soul. Of course they will go insane trying to stop it, and there will be a big debate about the property rights of "dead" people...