Sup /g/. I don't know if this is the right place for these questions and I'm sure I'll sound like an idiot since I don't understand the technology of this subject too well, but I wanted to talk about robotic prosthesis.
I've read in several places my whole life that "mind controlled" robotic limbs were in development/use, and I was under the impression that this meant that through some neural implant, electrical signals in the brain were transferred to a robotic prosthetic and basically it was as natural and immediate and smooth as moving a real limb since it was the same neural signals being sent, just instead of going to muscle it goes to motors and stuff.
However, I've recently been told by my medical school friend that almost all robotic limbs labelled "mind-controlled" are in fact controlled by electrodes which sit on the surface of the skin and active the prosthetic whenever the muscle they sit on twitches or flexes. He also said since the user had to think about what muscles to flex, then flex it, then wait for the controller to recognize it there was actually quite a bit of lag in these robotic limbs, and they could never be as fluid, or natural, or immediate as a real limb. Is this all true?
Since I can't find a lot of good information that I can understand on google, I just wanted to know if /g/ could provide some clarity on this subject. What are the different methods for robotic control of prosthesis? Can robotic prosthesis be truly controlled just by thought? Do they use neural implants/some kind of muscle implant for prosthesis? What are they pros and cons of the different methods? Is there lag in robotic limbs? Can they ever be as natural and seamless as a real limb? How much do they weigh? How long can they hold a charge?
Help please /g/, I'm absolutely fascinated by this subject and would like to learn as much as I can.
I'm a girl btw
bump
/g/ pls
>>52646244
Brain creates electromagnetic fields when neurons fire.
Watch for patterns in fields when people think about certain things.
Refine signals and program robot to react to certain signals.
Modern robotic prosthetics in a nutshell.
We're not at the point where we are actively converting signals from the nervous system
The nerves that travel from your spine to the muscle that flexes your hand/fingers etc still exist when you lose it, so what these "MC" limbs are about is reading the electrical impulses (detectable on the skin) from each of these nerves. Surgery can also be performed to move the nerve bundles to closer spots, hightening accuracy etc. There is a ted talk where one person "mind controls" another persons hand my having their nerves read, the other persons nerves were blocked with electricity or some shit and the first person's were relayed and emulated, causing person 2s hand to mimic person 1s. There are also 3d printed arduino based products that rely on this tech, so its not super advanced. The real advanced stuff is spinal implants that truly convert "thought to code" for lower body exo's
>>52646244
youve come to the wrong board
here we only post screenshots, homescreen, battlestation threads. as well as other high quality threads.
>>52647046
not OP but also very interesting in this subject.
does anyone know a good place to start?
>>52646449
that guy trying not to look at her
There's no neural implant.
Prosthetics that detect muscular movement are the most common. There are things that can pick up brain activity like an EEG and translate that into movement as well.
The user doesn't have to think "move arm" or something the same way you do not think about moving your arm before you do it.
https://www.youtube.com/watch?v=zVMDHCnT-d4
>>52647474
Can you blame him?
I work in an association concerned with the development of open-source medical devices such as prosthesis, ask me anything.
One of our projects is actually a muscle controlled forearm replacement, it's still WIP but it uses a couple of sensors to open/close the fingers on the arm.
>>52647965
What are the pros/cons of osseointegration when it comes to prosthesis?
>>52646244
Look here: http://www.cyberdyne.jp/english/products/HAL/index.html
>thinly veiled alizee thread
Thank god
>>52648969
Holy shit hhhnnnngggg
>>52648969
>>52648433
Osseointegration works only if there's at least some part of the bone intact (like, arm cut in half) but if the bone is completely missing up to its joint then OI is useless, also it needs the supervision of a medic; we don't do anything so sophisticated btw, for now we're using 3d-printable materials to print the prosthesises.
>>52647345
Look up Hugh Herr. He's got a Ted talk about this sort of thing and has done a lot of neat work with below the knee prosthetics.
As someone who wants to pursue this topic for graduate studies in a couple of years let me tell you what I know.
>I was under the impression that this meant that through some neural implant, electrical signals in the brain were transferred to a robotic prosthetic
That is indeed a certain kind of BCI currently under active development. The preferred method for this technique is known as ECoG (electrocorticography) and involves laying a grid of electrodes on the surface of the brain to measure its electrical activity. Here is an example:
https://www.youtube.com/watch?v=ogBX18maUiM
>it was as natural and immediate and smooth as moving a real limb since it was the same neural signals being sent, just instead of going to muscle it goes to motors and stuff
This is where we run into trouble. A single ECoG channel, though possessing a vastly better spatial resolution than EEG, still takes its recordings from a very large population of neurons. As such it receives a very noisy signal that are difficult to decode. This is an area of very active research.
The major drawback is invasiveness. ECoG requires surgery to install which immediately makes it the kind of thing that is inaccessible to most people. This is why simple EMG (electrodes on muscles) prosthetics are favored in a lot of cases, and really that's all you need sometimes:
https://www.youtube.com/watch?v=CDsNZJTWw0w
More Alizee and less autistic shit please
What should you be studying if you want to work with robotic limbs and other transhumanism like this?
>>52650663
It's a very multi-disciplinary area. You need biomedical guys, EE guys (or more specifically people that are good at signal processing), ME/CE guys for the robotics. If you're really curious look for some research teams and send them an email.
>>52650578
Not to mention cracking the skull open to get to the brain is highly invasive and neurosurgeons want to avoid that as much as possible
