This is a review of the paper “Chadwell, Alix, Laurence Kenney, David Howard, Robert T. Ssekitoleko, Brenda T. Nakandi, and John Head. “Evaluating reachable workspace and user control over prehensor aperture for a body-powered prosthesis.” IEEE Transactions on Neural Systems and Rehabilitation Engineering 28, no. 9 (2020): 2005-2014.” [1].

Discussion points:

• Correct shoulder anchor and cable mount technology appear to be missing from the study, and in fact, from most orthopedic practice.
  • A well-built[2] body-powered prosthetic arm for real work in my view is first and foremost a balancing tool, an asymmetry prevention tool, and an orthopedic means to reduce or prevent overuse, that also has to be used properly, and only a lower priority is assigned to make the wearer look like he/she has indeed a hand, when, in fact, she/he hasn’t.  The study under review here unfortunately does not offer any understanding or insight in the direction of understanding a prosthetic arm as a balancing tool and overuse prevention tool, which however should be the first priority.
  • The paper appears to employ a peculiar Figure-9 harness/cable mount, one that appears to be ill-conceived. In my own experience, a body-powered arm has to be built well [2] for real use, but also, before its analysis can deliver sensible information about this type of prosthesis [2]. Body-powered prosthetic arms when built by the current industries of prosthetic technicians supplemented by large manufacturers, all of which seem to have massive financial incentives to discourage body-powered arm use, tend to also exhibit faulty cable mounts, fault-prone wrists, and wrongly conceived harnessing with the maybe even intended goal of conveying a bad consumer experience. Fascinatingly, academic research joins them there, which may have to be the subject of some other dedicated study.
  • Poor mechanical efficiency points to poor body-powered arm-building skills. The authors submit that body-powered arms have “poor mechanical efficiency”. I mean, sure, if one lets a technically unskilled (that is: a person that also has not understood how Bowden cables and shoulder mounts should work, and do work) person put together some parts, then an unsatisfying device may be expected as likely if not guaranteed to come from that. The authors cited Hichert [3], where the study cohort consisted of people that (with 1 exception) were not at all habituated to body-powered arms, that apparently used a similar wrongly setup Figure-9/cable setup, to conclude that in essence, equipping people that never liked crappy body-powered arm builds with crappy body-powered arm builds would cause a bad experience. Modern user-based developments have left these considerable shortcomings far behind [2], but this approach does require an actual understanding of technical aspects.
  • Survival of the fittest. So as uncanny as it appears, the roles become thus reversed, in that some (but not all) people with disabilities manage to design, build and control technically better prosthetic solutions than industrial or academic R&D. That does not mean that efforts such as mine, to distribute and promote better technical understanding of body-powered prosthetic arm technology, are in vain or useless – they just do not arrive everywhere at the same time. Whatever you need for such is all linked or written on this website. For free.
  • Required bodily excursion to control the gripper of a body-powered arm can be reduced by a proper build. – There is an understanding to this, and it can also be described: an ill-designed build with a soft Figure-9 harness strap will, when pulling to open a VO device, first touch the skin, then over some further 4-5 cm of cable extension, compress soft tissues and brachial plexus but not open the VO device, only then get tension, and only over the ensuing next 4-5 cm of cable extension, actually open the device. So the arm may have to be extended over 8-14 cm before a VO device is fully opened. Using a proper shoulder brace, I get a full skin contact of a non-tensile brace, and only after some initial 4-7 cm arm extension, my device is fully open. The dynamics of proper cable housing and shoulder brace may be quite relevant there. It may get totally apparent if one wears it, see videos below. With a tight, well set-up control [2], the alleged disadvantages can be significantly reduced. There is a reason why my prosthetist calls me massively spoiled by good performance.
  • You also have to properly use body-powered control. This study documented how their “Ten healthy anatomically intact adults with no upper limb musculoskeletal injuries/abnormalities age (19-49)” had problems with fully opening their VO devices at close range. That can be a problem when the researchers and experiment instructors perform research on a topic that they do not know in too much detail about. Insufficient cable control in the close range usually stems from the user being unaware of extending the elbow to open the gripper all the way. Also, see the videos below.
• Correct posture is key for real work – and so it seems to not be a good idea to extend the arm fully, too often, for real work. The study correctly submits that “Where active control is required, upper-limb body-powered prostheses may be particularly suitable for those who undertake manual work and/or do not have access to reliable electricity supplies”. Reality has it, however, that most hazardous work – highly repetitive and/or heavy work – is not delivered at the outer range of extension of the disabled or intact arm. Correct posture particularly for manual work is a real thing! But you only know about it, if you really know about that type of work. So I pay permanent and close attention to working within the ideal range of wrist, elbow, and shoulder extension and flexion. There is a whole separate body of work about this, and you please go round it up yourself. If one considers that, then one will not assume that “a larger volume of reach” is automatically better. To go with such an assumption, however, implies that the authors of the study may have no actual experience in relation to real work. A likely own absence of authors from any extensive real work explains easily why their research results appear to be almost fully disjunct with the real requirements for a prosthetic arm from a real work viewpoint [2].

So really, the study should have stated that as a limitation.

The authors of this study [1] possibly wrongly assume, that a body-powered prosthetic arm – that is designed to be used for real work – or in fact any human arm, prosthetic or not, needs to extend towards an imagined maximum, far beyond what makes sense for real (hazardous level type) work. In reality, that is not the case, much rather, for real work, proper posture is different, whereas muscles are used in the range where they are most efficient. So that research premise does not seem to apply. Starting from that (wrong) assumption, the authors apparently used an insufficiently understood and poorly designed harness and cable mount design, which may best be described as a weak attempt at provisionally equipping someone for light activities, but, definitely not real work. The users that populated the study cohort appeared to be novice users that did not have any disability, and that therefore also had lacked any exposure or proper training, also to approximate a real-life use with these ill-devised prosthetic device attempts. In real life, a well-built body-powered arm will be able to easily cover the whole reach space with a full grasp range without issue, but in real life, also, a hazardous (repetitive/heavy) work use of a human or prosthetic arm will not occur at the outer margins of a reach space but within the range of safe loading.

And of course, on occasion, I also perform overhead work, and all sorts of extended arm work, that is only cool, because I do it rarely – and the reason I say that it is “cool because of only occurring rarely” is not because of prosthetics, but, because of ergonomics. Almost all hazardous work is done in a closer range, in a range with at least slightly flexed / bent joints, due to better muscle / postural efficiency, to protect shoulders and elbows, etc., so a maximal range is not required, particularly for real work. Having a maximal elbow or shoulder extension range at one’s disposal may be more an issue for dancers or for sports, where one extends an arm fully, but for different reasons/applications than “real / physically demanding labor/work” – but there, one typically wears a passive arm.

Demonstration of reachable workspace performance characteristic

When keeping the shoulder girdle still and fixed, I extend my arm only by 5 cm to go from 0 to max in terms of gripper excursion / VO device opening. I can cover for these 5 cms in real life by buckling my back, extending shoulders, and pushing my elbow out or forward while even keeping the device still / stationary. It will be part of subconsciously sucking up the best way to operate a well-built [2] body-powered arm to identify these control patterns all by yourself.

But it may help to, from time to time, tell others that they want to consider that before going out to tell others just how they managed to combine lack of experience, lack of knowledge with poor design / prosthetic build to arrive at conclusions that were not necessary.

I did optimize my prosthetic cable/shoulder mount problems for many reasons, but performance clearly was one of them. But I would never have considered analyzing reach if it was not for the question, why this paper [1] seems so interesting as it apparently combines the wrong things.

Hosmer 5 – demo of gripper single location precision

The first technical question, before the subject “reach space” becomes interesting, is how stationary my gripper is while I grip, so how precise is my control with regard to objects that are close (or even further away) from me.

Well, from my history I never guessed this was a “thing”, but it seems it is a thing as others may struggle a lot from the lesser inspired design attempts for body-powered prostheses that they get built.

So, this arm here is easily used to provide extremely stable grip localizations. There is no problem whatsoever, however, the technical basis is what I call a well-built [2] body-powered prosthesis. I would at this stage assume that there is quite simply no way around a very good sturdy and good engineering build. However, I assume that, after all, we did use some nifty techniques.

Hosmer 5 – reach space

This standard device should be operated with precision with regard to close/open status regardless of where within the reach playspace/workspace one is. I try to show that this is possible also for close to the body positions as well as for far up, far out, far down positions.

TRS Jaws XFS – reach space

Here, the TRS Jaws even has an elongated (!!) lever.

People that already have to extend their arm by 15-20 cm before their device opens may not want that.

Me on the other hand (broom ting) has sufficiently well designed cable control performance to muster even close or far range full grasp range, without issues.

Conclusion

The world is full of people that do not know how to build, or, use, a prosthetic arm.

Even the so-called experts usually seem to be far from actual design knowledge, and that may be a surprise.

Exactly that though is why I wrote a while back that arm amputation shares properties/aspects/issues, as a condition, with rare disease/orphan disease type conditions: they are so rare that (virtually) no one is able to understand the aspects of them. Worse: just because everyone thinks they SEE what (they think) is wrong, does not mean they can fix it. Just because people see a cable, does not mean they understand how the cable works. Leave alone, how to set one up properly. Just because they see a cable housing does not mean they understand it.

If ever you want to build a body-powered arm that, for heavy and repetitive works, alleviates unilateral intact arm-related loads and instead allows for bimanual load distribution, then, getting the cable right should be the first thing you consider doing. Of course, one may not know many things that are essential to getting things right. How to mount a steel cable properly so it does not bend or abrade/chafe? How to mount a steel cable so it lasts very, very long? How to mount a steel cable so it only generates 50-70% of the usual curved deflection hysteresis compared to the material onslaught brought on by conventional attempts? How to build a shoulder brace so the brachial plexus is not compressed?

This calls for a series of documentation that is so bare and basic, that people that are interested will have a chance to understand it better.

So the study is a representative reflection of a student-level appreciation of a subject that seems ill-understood by academics and that seems equally ill-understood by most prosthetic practitioners. The industries – where I got my own know-how on how to properly develop a good setup – have great know-how, but, they are bicycle tuning specialists, bicycle part manufacturers, and steel cable industries that equip elevator/lift companies and the likes – in other words people whose very vital core business interest it is that cables run and operate well.

One very logic consequence is that it becomes clear that it seems to be that there are people with absolutely no skin in the game, that seem to tend to the subject of prosthetic arms, and (… and you can now finish that sentence yourself). Or reversely put: it ist 2021, and by far the technically best thing to wear for real work is my body-powered arm, with a number of own developments / components on it. Where is  the rest of R&D, and why? And why will it become 2023 and my setup will still be the best for any real work? And 2030? What is really going on there, that it appears that people facing better answers seem stick to the design mistakes from the fifties – almost as if it was a religion?

The real answers about the tragedy of prosthetic arm design are not technical. They are necessarily social and societal, and they will have to focus on values, and interpersonal dynamics.

@article{chadwell2020evaluating,
  title={Evaluating reachable workspace and user control over prehensor aperture for a body-powered prosthesis},
  author={Chadwell, Alix and Kenney, Laurence and Howard, David and Ssekitoleko, Robert T and Nakandi, Brenda T and Head, John},
  journal={IEEE Transactions on Neural Systems and Rehabilitation Engineering},
  volume={28},
  number={9},
  pages={2005--2014},
  year={2020},
  publisher={IEEE}
}

@article{schweitzer2018casestudyprostheticarm,
  title={Case-study of a user-driven prosthetic arm design: bionic hand versus customized body-powered technology in a highly demanding work environment},
  author={Schweitzer, Wolf and Thali, Michael J and Egger, David},
  journal={Journal of Neuroengineering and Rehabilitation},
  volume={15},
  number={1},
  pages={1},
  year={2018},
  publisher={BioMed Central}
}

@article{hichert2018fatigue,
  title={Fatigue-free operation of most body-powered prostheses not feasible for majority of users with trans-radial deficiency},
  author={Hichert, Mona and Vardy, Alistair N and Plettenburg, Dick},
  journal={Prosthetics and orthotics international},
  volume={42},
  number={1},
  pages={84--92},
  year={2018},
  publisher={SAGE Publications Sage UK: London, England}
}