Why is so much effort going into sensory feedback type military research and not rehabilitation research?

• Myoelectric control is inherently and unfixably unreliable anyway in daily use by amputees - but possibly not so much of an issue when used by an anatomically intact individual in an army robot control room.
• Myoelectric grippers are caught between sufficiently light (and far too weak) and sufficiently strong (but too heavy) without way out, from view of a prosthetic arm wearing community.
• They are a niche product even in terms of actually sustainable prosthetic fitting, from view of applied usage.

Logically, all the research effort that significantly helps military applications, but does not significantly help a real prosthetic arm, clearly marches into one direction only: army development, military research.

And because no one looks and no one cares, research money for rehabilitation of amputees can easily be siphoned off for military applications.

The background for asking these questions is that there must be very distinct reasons why in 2018, a body powered hook is still the only prosthetic type that can be reliably used in strenuous physical applications.

And now, we are starting to get interested in the sociological reasons why that is.

A recently discovered surprisingly high degree of cynicism towards disabled people and particularly those with an amputation by those that claim to technically improve rehabilitation very clearly points towards a non-rehabilitative sociological setting, whereas cynicism in army circles is to be expected [1].

The figures that underlie these indications are clear.

For telerobot control of a prosthetic or robot hand,

• vibrotactile feedback essentially halved the error rate or reduced it by ~50% in one study [2]
• elsewhere it reduced errors even by ~70% (cited in [3]).

For myoelectric users, the gains are marginal at best, despite experimental surgery with temporary nerve implants, or, stationary computers, such precluding any long term real-life use with very reliable results in a heavy usage exposure setting anyway,

• with a performance time that was shorter when using vibrotactile feedback (93.2 ± 9.6 seconds) compared with the performance time measured when vibrotactile feedback was not available (107.8 ± 20.3 seconds) (reduction of  ~14%) according to one study [4], while
• most halfways useful improvements appear somewhere between 0% and marginal [5], and
• my own exposure to sensory feedback on the arm resulted in very problematic experiences really [link].

The development of vibrotactile feedback for prosthetic arms is neither viable, interesting, financially relevant or functionally overly important, which bears the question why people with a particularly high degree of cynicism follow that line of investigation.

The essential conclusion is that the potential and the benefit of developing sensory feedback for military and similar applications is huge, whereas it is (at best) marginal for experimental settings in amputees and entirely unrealistic for mobile long-term equipment of heavy physical exposure type application in amputees.

So, instead of improving the situation for reliable physically demanding tasks, amputee research is likely done as an excuse to use rehabilitation research resources for military and similar robot control application.

That also would explain the daft approach to prosthetic arm testing that is currently taken [link, link, link].

The underlying issue is that since decades, hard demanding work only is possible with a body powered hook. Research has promised and failed for decades. It is now time to shed light into the social and sociological reasons of possibilities why that might be. If decades of proclaimed, advertised and hyped up academic research leave the arm amputee standing in the rain, where he was standing before, then we have a right to put two and two together ourselves. If you do not want that path to be taken, you might have investigated better prosthetic arms and actually followed that line of investigation a decade or two ago. I am obviously not saying that advancing military research through the tricks of siphoning off disability funding, or using public attention to the direction, is a criminal act - I could not possibly comment on that. But I am saying that there are striking issues with disjunct choices of topics, and interesting aspects of performance (above) in the wider context that are definitely interesting.

@article{braithwaite1984cynicism,
  title={Cynicism amongst military police personnel in Western Europe},
  author={Braithwaite, Lloyd and Sonnad, Subhash R},
  journal={Justice Quarterly},
  volume={1},
  number={3},
  pages={413--436},
  year={1984},
  publisher={Taylor \& Francis}
}

@article{lathan2002effects,
  title={The effects of operator spatial perception and sensory feedback on human-robot teleoperation performance},
  author={Lathan, Corinna E and Tracey, Michael},
  journal={Presence: Teleoperators \& virtual environments},
  volume={11},
  number={4},
  pages={368--377},
  year={2002},
  publisher={MIT Press}
}

@article{chen2007human,
  title={Human performance issues and user interface design for teleoperated robots},
  author={Chen, Jessie YC and Haas, Ellen C and Barnes, Michael J},
  journal={IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews)},
  volume={37},
  number={6},
  pages={1231--1245},
  year={2007},
  publisher={IEEE}
}

@article{raveh2018evaluation,
  title={Evaluation of the effects of adding vibrotactile feedback to myoelectric prosthesis users on performance and visual attention in a dual-task paradigm},
  author={Raveh, Eitan and Friedman, Jason and Portnoy, Sigal},
  journal={Clinical rehabilitation},
  pages={0269215518774104},
  year={2018},
  publisher={SAGE Publications Sage UK: London, England}
}

@article{stephens2018review,
  title={A review of non-invasive sensory feedback methods for transradial prosthetic hands},
  author={Stephens-Fripp, Benjamin and Alici, Gursel and Mutlu, Rahim},
  journal={IEEE Access},
  volume={6},
  pages={6878--6899},
  year={2018},
  publisher={IEEE}
}