To have “biomimetic” properties is a thing for prosthetic arms these days.

In this context, the term is used synonymously with “anthropomorphic” [1]. While that paper offers a lucid moment by starting out with “anthropomorphism could mean different things to different people and therefore is a subjective criteria, influenced by the knowledge and experience of the user”, it then continues claiming their approach to be “objective”.

Clearly, we are dealing with a purely user and experience based type of dimension.

As I see some aspects of my own body-powered prosthesis as highly biomimetic, and as that is not reflected by current product advertising or academic papers, I explain how biomimetic and anthropomorphic are vague terms that are clearly relevant, in that they may also reveal more about your own way to be human.

Correlating biomimetic properties of prosthetic arms

As a whole range of properties that are arguably biomimetic or anthropomorphic do not go together in what really are highly constrained prosthetic arm designs, the following table gives an overview of what in my experience can be combined and what cannot. Thereby, you cannot at once have a useful and useless prosthetic arm, as a simplified example, whereas some biomimetic aspects are in line with a high functioning arm, others obstruct such a construction or build.

The right column adds up all correlation points, to show which biomimetic property is ranked highest in a competitive sum game.

By that, real time speed, absence of limb positioning effect as well as function past thermal influence, particular profuse sweating, should be rated as particularly biomimetic. Other biomimetic aspects such as a cheap myoelectric device as such, which some people regard as pinnacle of anthropomorphism these days, lose out, because they are incompatible with a whole range of other relevant biomimetic aspects such as real time speed or reliability when sweating and that, due to that, don’t play well.  Also when placing a preference on cheap myoelectric finger control as one of the device’s biomimetic virtues, too many other biomimetic aspects are thrown out the window with that design choice. That is, however, how designers are locked in, constrained and it is up to them how they want to proceed.

The most interesting read-out of that table are two aspects though. Heavy and repetitive work balancing to prevent asymmetry issues to back, neck, shoulders, is one of the hard medical reasons to prescribe or wear a prosthesis for: alone that criterium adds a favorable number of other in part related biomimetic aspects, all of which add a different flavor of “being human” than an obsessive focus on tentacle shape. Keeping stump skin intact isn’t a prominent biomimetic aspect – it is more a medical aspect, so if one wants to look at a prosthetic arm  as a medical complication, that is more a priority there. And yet, keeping the skin entirely clean of abrasions or bruises is a key element in being able to wear the prosthesis also tomorrow. As having a human hand today and also tomorrow is a very anthropomorphic concept, the timeline aspect is interesting also in this context.

And then there are bi0mimetic criteria that almost don’t touch the other ones, that stand alone. They can always be added, improved, without biomimetic cost, so to say. To make the whole prosthesis as such very durable (including all parts of it), to design the  gripper angle so less trunk compensation is necessary [link], and, easy to add soft grip covers for grip force enhancement [link], are often underutilized possibilities that do not preclude the other biomimetic aspects in that they turn out to be more or less independent of the other ones.

Temperature/sweat independent availability of reliable (R) function [BP]

As a human, I perceive being alive and active also – and to a fair degree – as being able to work, play, live, when I sweat [2].

From that book: “Perspiration is frustratingly out of our conscious control. You can hold back other inopportune bodily functions—tears, burps, farts, pee, poop—at least temporarily. Not so for sweat. When our core temperature rises, the information is dispatched unconsciously to our brain’s hypothalamus. That’s where the executive decision is made to activate the skin’s sweat glands, and there’s no amount of willpower that can stop this.”

Clearly a near 100% control success rate for prosthetic control isn’t asked too much in this day and age [link]. Myoelectric arms are not ideal for adults [link]. And not only physical activity, but just nervous stress, will easily start a sweat and myoelectric failure orgy [link].

So having a prosthetic arm that purely delivers, in terms of control and performance,  while sweating profusely, is extremely biomimetic.

Real time speed [BP]

My prosthesis being as fast as my body is a simple to understand human like, or anthropomorphic, or biomimetic, property. While myoelectric control has it difficult to even approximate anything close to that speed, its worth as a truly biomimetic aspect has been recognized and if only indirectly [3]: it is good engineering practice to build the main features in as main building / construction principles, so when I know that a real time speed sweat proof control is what I want, then I will not find myself the slowest and most skin humidity sensitive contraption I can find and then waste years tinkering with that. Then I’m a tie a mechanical cable on my arm, make that a super comfortable tight well dimensioned snug setup and take it from there.

As things are, a well built body powered prosthetic arm’s speed is clearly unparalleled [link].

Lack of superfluous weight due to light body powered Design [BP]

Center of gravity is always a big thing when wearing a prosthesis every day [link], as abrasions, bruises and blisters can be a consequence [link]: requiring me as amputee to suffer blisters for you as bystander to feel less inhuman by what you project on me reduces your humanity, makes you fail the Voight-Kampff test. So there goes your biomimicry with regard to lower weight.

High control reliability (R) [BP] and Absent limb positioning effect [BP]

As stated above, clearly,  a near 100% control success rate for prosthetic control isn’t asked too much in this day and age [link]. Myoelectric arms are not ideal for adults as we older people sweat too much [link]. And not only physical activity, but just nervous stress, will easily start a sweat and myoelectric failure orgy [link]. Also lets us not forget the limb positioning effects [link] which is a funny control problem that researchers often fail to understand due to their lack of knowledge of anatomy [link].

Also, the social stigma associated with loss of control generally, as evidenced in neurological disease and disability [4], will be probably similar when a prosthesis works with any error rate above 0.01% – so real life myoelectric control error rates in the 5-25% range are a definitive stigma, even though nondisabled bystanders may try to avoid to overtly admit that.

If you don’t believe me, drop your pen to the floor every fourth time you grab it in any meeting you have. Try out your colleagues’ reactions. You will find out.

So a near perfect control reliability is a biomimetic aspect that, if ignored, points to severe thinking problems in the affected engineer or researchers, just to get that out. We don’t tend to regularly wear stuff that does not work, and “work” really means that an industrial level of grip reliability is far from facultative or nice to have but terminally required [link].

Device WL real work robustness [BP]

The usefulness of a prosthetic arm for real work also depends on its robustness. There, the weakest link usually limits the fun that can be had.

As an example, I started out by being given a prosthetic arm with an Otto Bock hand, an Otto Bock wrist and the Otto Bock perlon cable control setup with a Figure 9 harness. The hand irreversibly locked within minutes that I had that prosthesis. The perlon cable irreversibly degloved/ tore shortly after that and also Otto Bock’s steel cables were mounted like a hot mess [link]. These then were changed / repaired at the next available appointment, like, two weeks later. The wrist wiggled a few weeks after that as its spring was too soft,  really a mechanical design not built for any longevity [link]. The hooks that I got after that hand had died were differently difficult, but “particularly robust” was not the property that came to my mind [link]. The Figure 9 harness caused me plexus compression problems [link]. And the strangely shaped liner by Ossur caused me skin problems [link]. So these parts, out of the Otto Bock’s body powered list of components and the Ossur liner, could be shown to be not robust and not up for real work.

We replaced all of these parts [link].

Anyone can replace all of their prosthesis’ parts – make the whole design actually fun to wear while durable!

As example, a myoelectric wrist does not have to combine mechanical and electric connection, if one really wants, one can make these separate.

Comfortable grip access angle [TD]

The reduction of compensatory trunk motion isn’t that hot as subject – unless I use the prosthesis to actually work,  i.e., to provide real work. Then, the heavy loads and repetitive loads [link] are applications where I like to see a good access / approach grip angle and as my options are limited for compensation, best to design the gripper shape with its opening and grip angle so less trunk compensation is necessary [link].

As Hosmer hook is a skeletonized and as that optimized shape embodying a thumb/index finger precision grip. It is a marvel of an engineering feat, unlike the grip angles enforced on the user by, say,  an iLimb, that lacks such considerations, and where, by that alone, no prior extended use/testing is plausible to ever have happened, not at the level of where Hosmer hooks were built and tested [link].

Cheap ubiquitous hand design thinking [TD]

The idea that any hand-like shape makes the user happy is, design/use wise, funny. As it is the opposite of the comfortable access angle. As that, this design principle negates, excludes, a few use cases.

Cheap ubiquitous myoelectric finger coordination control [MYO]

To just put motors into the device and let these activate without further coordination is a design decision that saves the designers time and money, and that saves the user extra parts/weight and cost. Or, a decision that, as in the case of the iLimb, increases the manufacturing cost and sales prices difference, to yield better income to Ossur or Touch Bionics.

The issue there is that finger tips do not always touch each other with precision [link]. Which will cause “precision grips” to systematically fail. I had alerted Touch Bionics and their “pilots” / “brand ambassadors” also of that issue, which did not prevent exactly these people from failing to win the 2016 Cybathlon, to a substantial degree because of that [link]. However, the cinematography of dramatically enacting human-like hand motion still occurs – which somewhat exposes such a device as theatric acting prop rather than a workman’s gripping device, whose main role may be to use it as placeholder for a real brand name to increase social status [link].

Myoelectric motors and battery with sufficient power [MYO]

These motors and batteries do try to provide power, that, however, I,  as myself (check blog title again? you good?) bring to the table as is. So I can “body” “powered”. Adding more power in terms of motors, and adding required batteries, fuel, etc.,. is simply redundant and thus unnecssary. Regardless of that, where to put that type of deadweight?

A reasonably powerful prosthetic hand with all superbly finely grained controls and powerful motors will add some 7 kg. Anything less is that: less. Nevertheless, the weight and weight distribution [link] in context of a typical myoelectric socket will cause, easily, a plethora of abrasions, bruises and blisters as a direct complication [link]. We have a Federal Law that says that if the prosthesis damages my health I am not required to wear it ¹.

Seeing as if “power” is a term that populates what people assume biomimicry to be, it is relevant to list it here [5].

Socket comfort with gel liner/cotton tube gauze and no electrodes [BP]

That one has been the most frequently asked question on any support forum, ever. So I am not going to explain this any further. It seems to be a fact that current prosthetic arm biomimeticists conveniently ignore that sustained perfect comfort is the quintessential requirement for body integration and thus, as such things don’t fall out of the blue sky, they could occupy themselves with that. Particularly with that. Like, to start there.

Read this: LINK.

Socket with electrodes and skin issue/burn risk [MYO]

Reverse reality is also used in biomimicry for any reason.

Read this: LINK.

Heavy /repetitive work is sustained, for balancing out asymmetry of body [BP]

Read this: LINK.

Easy to add soft padded grip enhancement (gloves, tape) [BP]

Read this: LINK.

Conclusion

It is up to you how you want to live, perform heavy and repetitive tasks, survive asymmetry. Are you an action figure or a barbie doll in human cosplay? Or if you want to embody robotic dreams. Many aspects define biomimicry, and wearing a highly functional and comfortable prosthetic arm in no way makes you less human than wearing a 7 kg robotic demo piece that makes you break out in sweat all the time. You decide if you want to invest in this or that, buy this or that. You decide if you want to test a device or go ahead and just buy it. Have it your way!

@inproceedings{kakoty2013biomimetic,
  title={A biomimetic similarity index for prosthetic hands},
  author={Kakoty, Nayan M and Hazarika, Shyamanta M},
  booktitle={2013 IEEE Symposium on Computational Intelligence in Rehabilitation and Assistive Technologies (CIRAT)},
  pages={32--39},
  year={2013},
  organization={IEEE}
}

@book{everts2021joy,
  title={The Joy of Sweat: The Strange Science of Perspiration},
  author={Everts, Sarah},
  year={2021},
  publisher={WW Norton \& Company}
}

@article{schone2024biomimetic,
  title={Biomimetic versus arbitrary motor control strategies for bionic hand skill learning},
  author={Schone, Hunter R and Udeozor, Malcolm and Moninghoff, Mae and Rispoli, Beth and Vandersea, James and Lock, Blair and Hargrove, Levi and Makin, Tamar R and Baker, Chris I},
  journal={Nature Human Behaviour},
  pages={1--16},
  year={2024},
  publisher={Nature Publishing Group UK London}
}

@article{warner2018factors,
  title={Factors affecting ability and satisfaction with social roles in persons with neurological conditions: The importance of mobility and stigma},
  author={Warner, Grace and Desrosiers, Johanne and Packer, Tanya and Stadnyk, Robin},
  journal={British Journal of Occupational Therapy},
  volume={81},
  number={4},
  pages={207--217},
  year={2018},
  publisher={SAGE Publications Sage UK: London, England}
}

@article{lepora2013state,
  title={The state of the art in biomimetics},
  author={Lepora, Nathan F and Verschure, Paul and Prescott, Tony J},
  journal={Bioinspiration \& Biomimetics},
  volume={8},
  number={1},
  pages={013001},
  year={2013},
  publisher={IOP Publishing}
}