The "Russian arm" control paradigm for myoelectric prostheses [still relabeled as "bionic" or "thought controlled" arm] {what we have been withholding from you so far}

The "Russian arm" [link] was introduced so many years ago. Then, it fulfilled a different role than just getting amputees back on track. Really, it was a modern version of the Carnes arm, that also was leveraged as somewhat political tool. And quite possibly, for finances.

Because, see, every age has its gadget arm.

Also, these (our) times have their "gadget arms". And their actual real shortcomings? They are ever so gloriously evaded, ignored, dismissed, forgotten, not mentioned and unknown.

About gadget arms

Appearances matter over everything else. So for the amputee deprived of friends, family and hope, and a hand and part of an arm of course, it just has to be the closest to a human hand, look-wise. It has to be the most modern, what is offered technology-wise. And among us, who really knows what is modern and not? So, media and hype are the ultimate hand makers. What they present and hail, that is what looks larger than what is: the actual gadget arm.

Every age has its gadget arm. And gadget arms are made, pushed, idealized, portrayed.

There is the client side.

Because, see, the hope runs high with amputees. Arm amputees every now and so often are close to clinical depression, and quite possibly, a lot closer than one thinks. Not the ones that don't give a damn. But those that go WTF, DAFUQ and Come On. If your insurance not only pays for prostheses but also for disability benefits or disability support, you can ask them about that. Call them, and find out if they rather pay for an expensive prosthesis, or, disability support for a deeply depressed arm amputee? They much rather keep the amputee happy - even if it takes them to finance a technically useless gadget arm. But dig deeper, on the phone with the insurance representative. And they might tell you that really they - as insurance -would be interested in durable, useful, powerful and reliable prosthetic arms. So, there go your unheard voices. A depressed and emotionally drained arm amputee, would that rather be a person to rage against not getting the "best" most "expensive" gadget arm, or, would that more be the person that diligently and patiently develops a better technical solution knowing that all true development takes grit, effort, brains, time and patience? How much mental problems do we need to believe that a "gadget" quick fix can "fix" us, or, at least, make us happy? What type of emotional disconnect is required to elevate these "fixes" to something great, superb or even glorious?

As, you see, there are deep reasons, why every time has their gadget arms.

Then there is the maker side.

Prosthetic hands and arms are not a massively profitable business, though. They are at best a show case for superb technology for the manufacturer. At best, they make the manufacturer shine, and look good. However, they only pay for themselves if they do not cost much to begin with. So there is your internal contradiction. In other words, if you look at commercial value, prosthetic hands and arm parts must be, quite necessarily, crap. If you want to be good at producing our time's best gadget hand, it should look fabulous, but the parts and inner parts do not have to live longer than 3 months of 2 hours of daily use - that is, approximately, 200 hours of play time. Because that seems to be how long the attention span lasts until the toy gets dropped. So you cannot take it personal that manufacturers will use quality control exempt bolts, el cheapo rechargable batteries, unreliable grip mechanics, and other questionable technical solutions, to stuff into these "gadget" arms.

Because as we were discussing, deep reasons exists that go into why every time does have their "gadget" arms.

So with reasons being reasons, having an insurance pay your "bionic" arm does not automatically make that arm useful or technically good. Not in the mildest way.

Carnes Arm, and Carnes Arm problems

And so in 1910 and after, the Carnes arm was of limited use also because of an explicit construction deficit. Its gears and string or cable control were so lossy or imprecisely manufactured, that employing its grip powers for 20 minutes would bathe the user in total sweat. The actual solution to this problem would have been a manufacturing process that allows for very high precision, or, strong lever arms, or, minimal amount of parts. Instead, the Carnes arm's mechanism does not employ strong lever arm mechanisms, the parts are far too numerous and any manufacturing to achieve low friction here will have to be massively precise. Even then, high part count also makes for a massive weight. Only a full understanding of these mechanical principles will allow a designer to fix the issues of the Carnes arm. From that time, the only mechanical hand that used the combination of  low part count, with low friction and weight, ideal strong lever arm mechanisms, and relatively good precision for manufacturing its mechanism, is the Becker hand. Its design is totally ingenious for the time. It still is unbeaten in terms of durability, weight, design, capability and price.

Russian Arm, and Russian Arm problems

In 1960 or around there, the Russian Arm introduced the two skin electrode method for myoelectric arm control [link]. With that, also a number of ystematic problems were introduced. Firstly, batteries would run out. Secondly, the socket had to always place electrodes on the skin. So soft materials were not an option, semi or fully rigid materials were the way to go. Thirdly, electric motors and control parts, hand parts and all made the devices heavy, so their centre of gravity was always far out distally, towards hand or wrist. Fourthly, co-activating opening or closing muscles with flexion of the elbow would make it virtually impossible to achieve a really reliable grip. It is a design feature of two-electrode myoelectric hands that you can neither lift a glass or cup to your mouth, elevate movie tickets to hand them to a cashier, or hang laundry. So fifthly, myoelectric arms are uncomfortable (way more than body powered sockets) due to socket design, they are unreliable due to electrode contact, battery, interference issues and co-activation issues that are inherent and unavoidable, and they have their center of gravity so far out that typing with a modern iLimb will hurt my shoulder after 10 minutes already. Ultimately, a number of outspoken gadget arm fans and wearers contracted serious overuse  problems of their (other) (human) arm and hand. They could have gotten that cheaper. Also, one can definitely anticipate these overuse problems using only thinking and reading skills [link][link].

So let us look at these issues more directly.

Socket comfort

As I posted already [link], stump pain prevails here, after wearing a modern gadget arm for not even a whole day. I am not the only one. You want to sell gadget arms, and succeed on the long run, fix that. If you ask me to wear that type of arm despite knowing this problem (view the graphical images! imagine wearing that for 4 months every day!) you will raise the question of how human you are. You really do.

That said, body powered arms also cause friction rashes and so on. They just do so with far higher loads and to a far lesser extent. After a few weeks you will notice the difference.

Myoelectric control error rates

In about four decades, academically published myoelectric arm control  error rates remained (a) unacceptably high and (b) became slightly worse [very detailed analysis here].

Unintended co-activation of myoelectric hand with elbow motion

Muscles that extend and flex the elbow also can trigger the myoelectric control for the terminal device. They systematically do so. And this is very important.

In the following two videos, I do NOT let go of the tickets, or cup, voluntarily.

The hand opens against my will simply by flexing the elbow, and it will remain closed against my will when I extend it (or open, that happened too). Simply, because that is how two electrode myoelectric control really works. It basically screws you. I cannot even reliably hang laundry with it.

Even though they probably mean well, even a recent assessment of control paradigms [1]  misses this real problem ever so completely. This is a major paradigm problem. How can that be? Simple: the authors write about something they never tried, and, they never made sure to understand by observation to its very bottom. The problem is far too common for them to miss it otherwise. So, that is unfortunate.

Image (C) Copyright Nicole Kelly

The following image shows how elbow extension clearly activates the hand's grip extension / opening, leading to the object (bottle) to fall down.This is not an electrode malfunction but a clear consequence of the dense anatomical build that we have, and that causes the rather cheap design approach ("place a myoelectrode here") to fail.

Here, an extended and lifted elbow with forearm extension prevents the myoelectrodes from a activating the close function. Another instance of limb positioning effect.

The advertising for Otto Bock Michelangelo massively targets "better posture" for their "bionic" hand. This is funny, because the general and overall posture with any prosthetic arm and hand is mostly perfect, while no actual grasp or holding action is performed. The problems invariably start when one goes about actual grasp and hold activities. The body powered arms require shoulder and elbow action to tense up the cable for opening (or, for VC devices, closing) action.

For holding on to an item once it is grabbed, a body powered arm will not require a particular posture. The myoelectric arm, however, does have the inherent capacity to royally screw you over - the hand may close (which is funny but irrelevant except when trying to get through a sleeve when putting on a jacket), but worse, open, if any of the muscles that extend or flex the elbow inadvertently activates the relevant electrode. With any elbow motion, the hand thus becomes functionally unreliable. As elbow extension and flexion then must be consciously suppressed, down the river goes your natural posture. So really, myoelectric arms are a total bitch inasmuch as posture is concerned - if you are actually using them for actual grasp and carry actions. With a few of these down the road, you will know better and from that point on, you will be totally reluctant to rely on a myo arm when holding something critical. In other words, you will overuse your other arm, or have posture problems, in a locked in situation that is design inherent with myoelectric arms. That is because electrode placement is optimized for sensitivity but not specificity.

An interesting attempt to solve this is a surgical procedure called "Targeted Reinnervation" [link]. There, unused nerve endings are spliced up and rerouted to surgically newly made muscle packs that the amputee then can activate. There are a few problems there, however, as well. Phantom and stump pain was anecdotically reported to be an issue. Moreover, if the then required precision for placing the socket mounted electrodes exactly over the muscle packs that were installed by the surgeon, the prosthetic device might not function at all - so there is no graceful degradation or partly available function but a total non-function. So the "Targeted Reinnervation" is interesting, but in no way does it simply blow away or resolve the socket and electrode placement issues (that are combined in what one might call a "hot mess").

Given that no one these days has solved the "Russian Arm" problems, they also cannot expect these hands to perform any better. Conversely, this takes off a lot of load off these hands: once these suckers do not excel in "function" anyway, why not keep them cheap, non-functional and moderately light weight? After all, we could argue all day why the general public agrees that "Gorilla arm" [link] strengths are not acceptable, so why would arm amputees suck it up and happily agree to a life with chronic bursitis, premature rotator cuff issues and so on?

These are several decade old unsolved relevant problems inherent to the myoelectric control. Solving them using, say, motion sensors or other electrode placements might be a path. Using only motion sensors to modify hand or grip, and body powered control for gripping, could be another approach.

But until these problems are recognized or even solved also by academia, this is not getting any better. Worse even - multi finger control is superbly capable if done in laboratory settings but not really under actual amputee conditions [2].

Until. These. Problems. Are. Solved.


With 80'000 CHF for such an arm, that under performs totally [cake link compare with body powered performance link] [car wash link compare with body powered performance link] [precision grip link compare with body powered performance link], performance alone is not the key to understanding its assumed "success" [or, see SHAP Southampton Hand Assessment Procedure comparison link]. Socket pain and skin damage are sizeable massive and stable unsolved problems [link]. Unaddressed overuse issues are another untapped big problem here [link][link].

With that not being problems enough, co-activation explains, in part, a very systematic repeated failure of these myoelectric arms. And while you can argue pressure blisters being a no go all day, functional failure - at dinner, at a restaurant, at a party, at work, in emergency situations - is not negotiable. Electrode malfunction, skin contact problems and myoelectric control issues like co-activation due to elbow function causing hand opening will be the most sustainable technical reasons for users to stop using their "gadget arms". I know that for me, that is a relevant aspect. Two other users that I know gave up their myo arms due to such functional reliability problems. They now solve their bimanual tasks without any prosthetic arm, and have to ill effects because of that.

Given the combined total distress that these issues inflict on the user, including very frequent costly repairs, any normal person will totally understand why myoelectric arms are a promise, a hope, an attempt maybe, a Dead Horse certainly - but never an actually sustainable solution.They are a Dead Horse because these issues have been known for a very long time, and so far, researchers and industrial developers found excuse after excuse not to solve them. There is a deep reason why every time still has their "gadget" arms. Mostly, we and our society have failed to verbally label and tag these individuals, the ones that get away with the sub-mediocrity of their "gadget arm" business model.

Even if these "gadget arms" are cool for the moment, they are truly useless for true bike riding, for hedge cutting that really goes down with the physical aspects and that still helps avoid overuse.

Now for something entirely different: this, here, is a Becker hand -  not myoelectric but body-powered, adaptive grip, very robust and affordable:

[1] Unknown bibtex entry with key [bonger2012bernstein]
[2] C. Cipriani, M. Controzzi, G. Kanitz, and R. Sassu, "The Effects of Weight and Inertia of the Prosthesis on the Sensitivity of Electromyographic Pattern Recognition in Relax State," JPO: Journal of Prosthetics and Orthotics, vol. 24, iss. 2, pp. 86-92, 2012.
  title={The Effects of Weight and Inertia of the Prosthesis on the Sensitivity of Electromyographic Pattern Recognition in Relax State},
  author={Cipriani, Christian and Controzzi, Marco and Kanitz, Gunter and Sassu, Rossella},
  journal={JPO: Journal of Prosthetics and Orthotics},

Cite this article:
Wolf Schweitzer: - The "Russian arm" control paradigm for myoelectric prostheses [still relabeled as "bionic" or "thought controlled" arm] {what we have been withholding from you so far}; published 01/06/2014, 13:38; URL:

BibTeX: @MISC{schweitzer_wolf_1656511962, author = {Wolf Schweitzer}, title = {{ - The "Russian arm" control paradigm for myoelectric prostheses [still relabeled as "bionic" or "thought controlled" arm] {what we have been withholding from you so far}}}, month = {June}, year = {2014}, url = {} }