It is well known among the few of us amputees that try to have a good excuse ready, that “lack of sensory feedback” never was a reason not to wear a body-powered arm.

Sensory feedback seems to be the new buzzword for prosthetic arm research. Now, the extent of just what a well-built body powered prosthesis performs also in this apparent domain of emerging new exploratory technology only became more apparent: the 2020 Cybathlon Prosthetic Arm race was won, as the 2016 Cybathlon Prosthetic Arm race, by a person wearing a body-powered prosthesis, with no added schnick-schnack. And that is interesting because the 2020 Cybathlon Prosthetic Arm race had a specific sensory feedback test part. It is maybe a factor that the creators of Cybathlon most likely did not know that body-powered arms, when worn by a below elbow amputee in particular, are relatively good there to begin with.

The thing is that, anticipating lengthy explanations and to put it bluntly, myoelectric prostheses are basically junk to begin with(link)¹, and to make these poor thingies at least a bit more accepted, R&D are deep into riding dead horses and adding features, furthering the already fatal death cycle more [link].

Now, we seem to deal with a particular legend, an urban story of sorts, if not a large-scale misunderstanding: as amputees that wear a prosthesis usually all totally depend on prosthetists in so many ways, these prosthetists sometimes also made retrospective studies. Also, they usually may have felt a bit sensitive about the prosthetic limbs that they had built, if not extremely sensitive, while what they had built really were usually rather fragile or breakable items, to use some euphemisms, nothing too useful, we are talking about prosthetic arms here, particularly myoelectric ones, and so without surprise, many amputees did not wear their prostheses too often, and instead of telling it how it is, as we all know how it is so what could possibly be gained from a truthful answer, one tells the prosthetist when faced with such a study question (“please answer honestly, no we can take it, sure”, lips trembling), one thus says very politely “you know I was not wearing the prosthesis, because I cannot feel just as well with it”, thinking that would put the possibly deeply disturbed minds of the poor tinkerers at rest. But then, they sat down and wrote the results up, and got them published who knows where, and now, a number of retrospective studies now list “absence of sensory feedback” as “reason” why amputees reject prosthetic arms. If ever they knew. It was never meant to end like that, but R&D mostly never really looks at what we really wear, they seem to have no idea that when one wears a body-powered arm with a hard wrist coupling, like I do, as most others, we have some sensory feedback, given restricted options though, but enough to get by in a way. Not enough so that a blind person could wear a prosthetic hook with same success though. So you understand: my stump is a bit of a jelly pain cake, it does not at all have a precise 2-point discrimination like a finger tip usually has. But it still amplifies even subtle socket-stump shape discrepancies relatively well, in a way that I think of my stump as an ‘easily triggered potato’ of sorts. So my prosthesis fits with a tight socket, and I feel some touch, and some vibration, quite well. Sufficient obviously to do all sorts of stuff, who knows if that bit of sensory feedback helps to reduce phantom pain even as well, I personally like to believe it is the compression [link] as the socket is a bit tight, not the feedback, but that is another story – at any rate, no need to immediately roll in the engineers to add electro-buzzers anywhere on my prosthesis.

So Cybathlon 2020 Prosthetic Arm race is over and Krunoslav Mihic wearing an apparently ~30$ 3D-printed plastic device carefully designed by Andrej Dukic of Croatia and apparently around just ~40 h of training won.

Now, I wonder whether I can show the extent of some sensory feedback in my normal body-powered prosthetic socket to be there, in a way that is easy to copy and try out as well.

For that, I used a new Powerball, and a Dynabee gyroscopic device [Mishler, A. (1973). U.S. Patent No. 3,726,146. Washington, DC: U.S. Patent and Trademark Office; link][1]. To some, it may be of relevance that the housing is (already?) colored blue [link].

This type of gyroscope was, and still is, sold as a toy or training device. Some of its technical gyroscopic aspects were first described, after we all had played with it, when I was a kid, and I do remember drawing and inking that hand in that paper there [Gerhard Schweitzer (1982) Antrieb eines Spielkreisels durch Taumelbewegungen seines Gehäuses, in: Festschrift zum 70.Geburtstag von Herrn Prof.Dr.rer.nat.K.Magnus, link][2]. Later, more technical articles appeared, one nice article describes also how a stable equilibrium depends on the phase angle [Gulick, D. W., & O’Reilly, O. M. (2000). On the Dynamics of the Dynabee. J. Appl. Mech., 67(2), 321-325; link][3].

The difference between my two depicted devices seems to be that the original Dynabee feels like it has heavier eccentric weights and thus feels like it gives back more momentum and resonating vibration than the newer Powerball, that instead also houses LED lighting. Both are relatively easy to start by pulling the rotor across some 20-30 cm of fabric (large mouse pad, trousers, ..) but to start the rotation off, very subtle momentum has to be employed, which I find impossible at this moment to do with a prosthetic hand.

However, once the device rotates, I am able to keep the older classic model running when then holding it with a prosthetic gripper that is able to provide an encircling power grip (I used a Becker Imperial hand, by Becker Mechanical Hands, with a PVC glove over it). The newer gyroscope that does not feel to me like it gives nearly as much vibration feedback, no way I could keep that spinning with the prosthesis, there I do not get enough very subtle force transmission, probably due to lack of too much neural real estate.

Video (below): For demonstration of how subtle momentum and vibration feedback is available in a normal body-powered prosthetic arm, I first start two different gyroscopes with my left (human, own) hand and later hand the spinning device over to my prosthetic hand (here: Becker Imperial Hand, PVC glove by Centri / Fillauer, Sweden) and the trick is to basically keep the gyroscope running by exerting a tumbling type of force against the device, which one may believe can only be identified correctly by feeling where the device wants to go to. On top, however, one also looks and hears but how to disambiguate this best? So this seems to be a reasonably good sensory feedback proof of concept test, whereas feeling with the arm / hand used seems to play a relevant role (but not proven), using a device that is cheaply available in qualities so similar I guess it would make for a cool test item.

The means by which the gyroscope is kept running is that I can feel the force and vibration that it exerts on my hand.

From the original patent description[1]: “The opposing or resisting torque referred to can be sensed by the operator and he soon learns to gyrate the supporting structure so as to keep the opposing torque at a substantially constant maximum value. (…) . In any event, the operator soon learns to control the gyration of the support structure of the device in accordance with the resisting forces he feels so that the rotor attains a high speed of rotation about its spin axis. “

At the same time, I am not entirely sure I do not also use hearing and seeing the thing turn for operation, which seem to be seamlessly overlapping with the feeling in the arm, which however *feel* like real. Also there is a distinct impression from the newer Powerball, that the reduced momentum feedback makes it almost impossible to successfully operate this gyroscope despite present hearing and vision.

While I am sure this should appear mostly useless for practical work with a prosthetic hand right now to most people from the other planet/s, these gyroscopes are both standardized and cheap, and so, why not use these for the next competition, just to see who can build the device with the better tumbling wrist motion and subtle / fine sensory feedback motion. This certainly will be increased to a maximum difficulty by switching off vision and hearing as other feedback of how the gyroscope is doing, but as I tried to explain, I feel as if the arm forces are a key element in successful manual gyroscope operation. Better, though, forget about it. Reading about some random guy that can keep a gyroscope running with a body-powered arm, right after a body-powered arm won the Cybathlon Prosthetic Arm race for the second time, i.e., in 2020, after 2016 was already also won by a body-powered arm, seeing as if the organizers had done their darnedest to make myoelectric technology “shine”, may just not be the right news for 2021.

But then, it just may!  After all, the Cybathlon’s very first and more unabashed, raw rules included a statement that “the rules of the competition should be designed in such way that the design of the novel technology gives the pilot an advantage over a pilot that would theoretically use a comparable but less advanced or conventional assistive technology”; seeing as if body-powered split-hooks are a far more novel and modern design than any hand shape worn for prosthetic outfit, even though it is not acknowledged by many in what appears to still be our partly barbarian societies (link), that rule’s intention obviously clearly went to directly, straight, immediately, identify the historically more novel and technically more advanced if not also more adequate technology by allowing it to win. Twice.

@misc{mishler1973gyroscopic,
  title={Gyroscopic device},
  author={Mishler, A},
  year={1973},
  publisher={Google Patents},
  note={US Patent 3,726,146}
}

@InBook{schweitzerdynabee,
title={Festschrift zum 70.Geburtstag von Herrn Prof.Dr.rer.nat.K.Magnus},
  editor={Schiehlen, Ulbrich, Schweitzer, Mueller, et al},
  year={1982},
  publisher={TU Muenchen},
author = {Gerhard Schweitzer},
chapter = {{Antrieb eines Spielkreisels durch Taumelbewegungen seines Gehaeuses}},
OPTpages = {83},
}

@article{gulick2000dynamics,
  title={On the Dynamics of the Dynabee},
  author={Gulick, DW and O’Reilly, OM},
  journal={J. Appl. Mech.},
  volume={67},
  number={2},
  pages={321--325},
  year={2000}
}