15 years ago the cost of something like that would be about $50k to $100k ... Literally costing you an "Arm" or a "Leg".
Look like JUST the hand is going to run you about $11k ... that doesn't include the rest of the arm + electronics, etc. The level of amputation also makes a BIG difference to the cost.
Now they just need to make it look more like a normal arm/hand!! I must say, I am impressed. This should be bought and paid for by the VA for any of our troops that need them!
The UI for that hand reminds me of how combos are done in fighting games! Way to stretch the heck out of two inputs. Any mention of "ticklers" for force feedback?
Now they just need to make it look more like a normal arm/hand!! I must say, I am impressed. This should be bought and paid for by the VA for any of our troops that need them!
How to go from a "one-off" R&D project to a mass production where they are cheap enough so every amputee gets one? Kickstarter? This would as big a leap forward as the arm itself
prof_braino,
"How to go from a "one-off" R&D project to a mass production where they are cheap enough so every amputee gets one? Kickstarter? This would as big a leap forward as the arm itself "
The "Arm to Human interface" needs to be a custom fit, it's not like buying a pair of shoes where one size fits several.
This custom fit involves ...
1)casting of the patient to create a negative mold
2)filling of the mold to form a positive form
3)contour reshaping of the positive form. This process usually highlights muscle contours and promotes a better fitting prosthesis.
4)vacuum form plastic (usually a skin soft derivative of Poly Propylene) over the positive mold to form a negative interface (aka socket) that the patient can directly wear.
5) And then you have a cosmetic shell that goes over the socket which can range greatly and further dictate the final cost.
Note: prior to any casting and depending on when the amputation occurred, there is a period of muscle atrophy and swelling reduction that must take place before any cast will be able to fit properly. This period can vary and sometimes requires "reduction socks" often used for diabetics to manage limb swelling ... i.e. ankles
If I ever needed a prosthetic arm I'd be asking for most of a swiss-army knife, socket drive, torch, blue-flame lighter, laser point, camera, etc etc and give the device a screen too. Possibly
android based controller to tie it all together and allow tinkering. And I'd program it to play Bach (because I can't).
A 3D scan is 'not good' as flesh deforms easily. You actually NEED to do a proper form as it will simulate the pressure caused by the prosthesis later.
A lot of that arm can be mass produced, but yeah, it's the interface and socket that will be the expensive part.
The reason that arm is only on loan is that it's a prototype. It's not necessarily built to the engineering standards required for it to have a long functional life. And they most probably are already working on the next generation prosthesis, while receiving feedback from him and possibly other users. (I'm guessing that they'll be working on noise-reduction, for one thing. That thing is too noisy. They may also be looking at some way of motorising the relocation of the thumb.)
I'm also assuming that when they get it back it will be completely dismantled in order to find signs of abnormal wear.
The muscle-interface is probably pressure-activated.
There's a lot of research going on into hooking up to muscle-control nerves and brain-signals(the guy who walked all those stairs using 'mind control' comes to mind)
There's even scientists working on direct camera connection to the brain/optical nerves in order to give blind people vision.
I believe that in half a decade, that prosthesis will look almost obsolete.
A lot of that arm can be mass produced, but yeah, it's the interface and socket that will be the expensive part.
Sounds like a good reason to have a set of standard interfaces between the socket and prosthetic. I.e. get one custom socket made and snap it into the latest "iArmX" once it comes out. Or conversely, when a new version of the human machine interface comes out, it can be snapped into old prosthetics. The latter use case is imho more likely, as the human machine interface is the hardest part and it's changing the fastest right now.
I think the recent progress at obsoleting the hook is due to two things. First, the recent progress on human machine interfaces. I.e. Myoelectric sensors (muscle activity) are shipping tech right now. While tech like the Braingate or the nerve re-location surgeries (i.e. This limb) are making "natural" control likely. Second, the Iraq and Afghanistan wars produced a large number of new people who want Luke Skywalker's replacement hand.
A 3D scan is 'not good' as flesh deforms easily. You actually NEED to do a proper form as it will simulate the pressure caused by the prosthesis later.
..... in half a decade, that prosthesis will look almost obsolete.
A 3D scan is a great place to start. Anything to get a jump on the manual process. The next trick would be automation to start casting a socket. IF it requires the subject to be present, fine. If it doesn't last 10 years, fine. Until somebody figures produces a "Mr Prosthesis", I'm thinking a person missing an arm would like to play with a prototype. In 6 years, the cost will go down so everybody can have one. I want two! (just for spares).
Just because its impossible is no reason not to do it.
I think we should figure out a kickstarter, so OBC can offer them on propellerpowered. Even if amputees don't like them, I have tons of robots that need a hand.
Not to throw water, but a 3D scan doesn't capture the proper contour that needs to be focused on... i.e. gravity would not be your friend in this case where 'holding it in place" with a casting procedure is closer to where you want to be rather than a 'free form" that a 3D scan would acquire.
As far as myoelectric sensors ... they haven't changes much in over 20 years. One of the leading suppliers for these sensors is a German company called OTTO-BOCK. There are other companies, but the preferred sensor, at least when I was in the field was from OTTO-BOCK as they were superior to work with as far as the signal to noise ratio.
The commands from a pair of myo-electric sensors have not changed either... Simply one muscle contraction causes change in one direction, while an opposing muscle contraction causes change in another direction. This change can be Open/Close , Left/Right, Up/Down ... a Co-Contraction of both muscles switches modes from Hand to Wrist, to Elbow, and in one case I wrote code to add a fourth function of shoulder. <- Lots of Basic Stamp I's and Basic Stamp II's reading the myoelectric sensors and provideing motor control associated with each function :-)
erco
Posts: 20,259
Comments
That's not the kind of loan that I would want to ever give back.. Wonder what the cost is?
Jeff
Look like JUST the hand is going to run you about $11k ... that doesn't include the rest of the arm + electronics, etc. The level of amputation also makes a BIG difference to the cost.
http://singularityhub.com/2010/06/30/how-much-is-the-newest-advanced-artificial-hand-11000-usd-video/
Agreed, though it's the taxpayers who would pay the bill. Certainly a small price for their sacrifice.
Yea, But I would rather my tax dollar was spent on that as opposed to sending my money to a foreign country that burns my flag!!!!
Lawson
How to go from a "one-off" R&D project to a mass production where they are cheap enough so every amputee gets one? Kickstarter? This would as big a leap forward as the arm itself
"How to go from a "one-off" R&D project to a mass production where they are cheap enough so every amputee gets one? Kickstarter? This would as big a leap forward as the arm itself "
The "Arm to Human interface" needs to be a custom fit, it's not like buying a pair of shoes where one size fits several.
This custom fit involves ...
1)casting of the patient to create a negative mold
2)filling of the mold to form a positive form
3)contour reshaping of the positive form. This process usually highlights muscle contours and promotes a better fitting prosthesis.
4)vacuum form plastic (usually a skin soft derivative of Poly Propylene) over the positive mold to form a negative interface (aka socket) that the patient can directly wear.
5) And then you have a cosmetic shell that goes over the socket which can range greatly and further dictate the final cost.
Note: prior to any casting and depending on when the amputation occurred, there is a period of muscle atrophy and swelling reduction that must take place before any cast will be able to fit properly. This period can vary and sometimes requires "reduction socks" often used for diabetics to manage limb swelling ... i.e. ankles
Sadly, I also know Beau's background and he's probably dead-on correct. (Thanks for the bucket of water there Beau!)
Didn't I read somewhere recently about a couple college students who created a universal arm? (very simple compared to this.)
android based controller to tie it all together and allow tinkering. And I'd program it to play Bach (because I can't).
Bucket of water? No.
Engineering requirements? Maybe a start.
As beau points out, the trick is the custom fit. This is not a show stopper, merely the engineering problem to solve.
Whats called for is a "fitting machine": stick you stump in here, and the machine (at least starts to) create the custom socket.
I think we could solve this. Particularly is we start to design now for tech that will be available in 5 years.
A lot of that arm can be mass produced, but yeah, it's the interface and socket that will be the expensive part.
The reason that arm is only on loan is that it's a prototype. It's not necessarily built to the engineering standards required for it to have a long functional life. And they most probably are already working on the next generation prosthesis, while receiving feedback from him and possibly other users. (I'm guessing that they'll be working on noise-reduction, for one thing. That thing is too noisy. They may also be looking at some way of motorising the relocation of the thumb.)
I'm also assuming that when they get it back it will be completely dismantled in order to find signs of abnormal wear.
The muscle-interface is probably pressure-activated.
There's a lot of research going on into hooking up to muscle-control nerves and brain-signals(the guy who walked all those stairs using 'mind control' comes to mind)
There's even scientists working on direct camera connection to the brain/optical nerves in order to give blind people vision.
I believe that in half a decade, that prosthesis will look almost obsolete.
Sounds like a good reason to have a set of standard interfaces between the socket and prosthetic. I.e. get one custom socket made and snap it into the latest "iArmX" once it comes out. Or conversely, when a new version of the human machine interface comes out, it can be snapped into old prosthetics. The latter use case is imho more likely, as the human machine interface is the hardest part and it's changing the fastest right now.
Lawson
I think the recent progress at obsoleting the hook is due to two things. First, the recent progress on human machine interfaces. I.e. Myoelectric sensors (muscle activity) are shipping tech right now. While tech like the Braingate or the nerve re-location surgeries (i.e. This limb) are making "natural" control likely. Second, the Iraq and Afghanistan wars produced a large number of new people who want Luke Skywalker's replacement hand.
Lawson
A 3D scan is a great place to start. Anything to get a jump on the manual process. The next trick would be automation to start casting a socket. IF it requires the subject to be present, fine. If it doesn't last 10 years, fine. Until somebody figures produces a "Mr Prosthesis", I'm thinking a person missing an arm would like to play with a prototype. In 6 years, the cost will go down so everybody can have one. I want two! (just for spares).
Just because its impossible is no reason not to do it.
I think we should figure out a kickstarter, so OBC can offer them on propellerpowered. Even if amputees don't like them, I have tons of robots that need a hand.
CAT-CAM (Contoured Adducted Trochanteric-Controlled Alignment Method):
http://www.oandplibrary.org/cpo/1985_04_015.asp
Fitting of a prosthesis:
http://scottsabolich.com/fitting/
As far as myoelectric sensors ... they haven't changes much in over 20 years. One of the leading suppliers for these sensors is a German company called OTTO-BOCK. There are other companies, but the preferred sensor, at least when I was in the field was from OTTO-BOCK as they were superior to work with as far as the signal to noise ratio.
The commands from a pair of myo-electric sensors have not changed either... Simply one muscle contraction causes change in one direction, while an opposing muscle contraction causes change in another direction. This change can be Open/Close , Left/Right, Up/Down ... a Co-Contraction of both muscles switches modes from Hand to Wrist, to Elbow, and in one case I wrote code to add a fourth function of shoulder. <- Lots of Basic Stamp I's and Basic Stamp II's reading the myoelectric sensors and provideing motor control associated with each function :-)