BSM Controller Similar to Controlling Natural Hand

The biomimetic sliding mode controller produced better results in terms
of position error, force tracking error and percent overshoot and was rated to
be more similar to the control of participants’ natural hands, when
compared with a proportional derivative force controller, according to a study.

“Design and control of a versatile upper limb prosthesis is a very
challenging task. While many breakthroughs have been made over the last several
decades, the difference in performance and quality between human hands and
artificial hands is substantial,” Erik D. Engeberg, PhD, of the
mechanical engineering and biomedical engineering departments at the University
of Akron, said. The study described the “development of a biologically
inspired application of sliding mode control and electromyogram signal
interpretation algorithms so that prosthetic hands may be controlled in a more
natural feeling and physiologically expected manner.”

BSM vs. PD

One amputee and nine able-bodied participants compared a proportional
derivative (PD) force controller with a biomimetic sliding mode (BSM)
controller, designed to map human muscle signals into prosthesis motor command
signals. Both controllers were evaluated analytically and subjectively through
position and force control experiments. Non-amputee participants used a Motion
Control Hand, whereas the amputee participants used his Ottobock SensorHand
Speed. After the experiments, participants rated how similar the two prostheses
controllers performed relative to the function of their natural hand.

Overall, Engeberg found that the BSM controller replicated the posture
of the human hand more accurately vs. the PD force controller. Although the
dual object lifting task was nearly impossible to perform with the PD force
controller, 89% of the participants were able to perform the task with the BSM
controller. Participants also had significantly less overshoot in the applied
grip force tasks with the BSM controller. At the conclusion of the tasks,
participants rated the BSM controller as good or better than the PD controller
when it came to controlling it in a manner more similar to the human hand
relative to position and force control.

“The position of the human hand is mapped to the prosthetic hand in
a more physiologically expected way,” Engeberg told O&P Business
“Also, the controlled force being done through co-contraction of
muscles actually is more similar to the way that force is controlled with the
human hand.”

Future studies, clinical practice

Currently, Engeberg is continuing this study of BSM control with a
dexterous artificial hand with applications for a new arm to be released by
Ottobock. Similar to natural systems, the BSM controller automatically closes
the hand for the operator, freeing an electromyogram control channel that could
be used for specifying the desired action of a different degree of freedom of a
dexterous prosthetic hand, for example, wrist motion.

“What I have is a dexterous artificial hand that has flexion and
extension of the wrists, as well as controlled force of the fingers,” he
said. “So extrapolating from the BSM control idea enables simultaneous
control of a wrist and hand. Ottobock is developing a new arm with a motorized
wrist and also the force control with the fingers. I have done some preliminary
studies with a couple more amputees with control of the wrist and hand
simultaneously and it looks promising.”

When it comes to clinical practice, Engeberg hopes his research provides
an easier training process for patients.

“What I hope will result from this is perhaps a shorter training
time when amputees are initially fitted for a prosthesis, so they won’t
have to think as hard about how to control the artificial hand for as
long,” he said. “What I’m most excited about is that this opens
the door to do simultaneous control of the wrist and hand at once. It’s
laying the foundation for more sophisticated controls.” — by Casey

For more information:

Engeberg ED. A physiological basis for control of a prosthetic hand.
Biomed Signal Process Control. 2012. [Epub ahead of print]

Disclosure: Engeberg has no relevant financial disclosures.

Leave a Reply

Your email address will not be published.