Page 183 - Handbook of Biomechatronics

Upper-Limb Prosthetic Devices 181

is going to dictate what the expected set of movements is for that person.
Of course, everyone would desire to have a hand or arm like before, delicate,
“flexible” which can perform almost everything. But reality is that technol-
ogy cannot do that nowadays. Therefore, as designers, we have to think
what is the priority for the person who is going to use that hand or arm.
What are the tasks that are mostly performed every day? That is, when per-
sonalization comes into the picture later on.
The expected set of movements has to do with how many-DoF the pros-
thesis can do. This has to do with the level of amputation and the needs of
the amputee. We have to admit that, the current way of how we manage the
prosthesis process, is not asking these questions and more importantly id does
not provide to the amputee a prosthesis that is fulfilling his or her personal
needs, that is, there is no personalization.

Control Method
As mentioned in Section 1.2, low mental loading which is achieved by sub-
conscious control is a key success factor for upper-limb prostheses. But how
subconscious control could be achieved? Subconscious control could be
achieved if the prosthetic action is integrated with sensory pathways which
inform the user of the state of performance of the task subconsciously.
The control method of many-DoF prostheses, should also be subcon-
scious, actually because of the many-DoFs, controlling many-DoFs simul-
taneously requires and demands even stronger the control to be
subconscious. If not the delays and lack of performance will be augmented
and will lead to poor performance.

Performance
As mentioned in Section 1.2, a key factor for success of prosthesis control is
the fast response of the controller. Natural delays in humans during reaching
from onset of neural command to execution of the task are in the order of a
few hundreds of milliseconds (200–300ms). Therefore, any delays in pros-
thesis control should not surpass these limits.
Other performance parameters are the weight and enough power for
1day of regular use (Bertos, 1999).
Higher usability should always be measured. In the past, it was shown
that high classification accuracy of many-DoF prostheses does not mean
higher usability (Lock et al., 2005).