You are on page 1of 16

Actuators in biomimetic prosthetics

Joey Leonard
Independent Research
25 April 2016
Advisors: William Aldrich, Raymond Gerstner, Nicholas Formica
Instructor: E. Leila Chawkat

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 1
Abstract
This research paper focuses on biomimetic prosthetics, and pertains to the recent
advancements made in this field as well as what improvements could be made for them in the
future. It reviews and analyzes journals, papers, and experiments about robotic prosthetic
prototypes published by accredited engineers and uses them to create original results and
conclusions about the topic. The purpose of this research is to show what role actuators play in
creating a robotic prosthetic that can mimic human movement and how to best create a product
that will effectively reproduce these movements, as well as highlight advancements made in a
field where they are much needed. Through the research, a relationship between the complexity
of the prosthetic arm and the ease with which an operator can use it at a high success rate is
established. After observing results and data about current prototypes being tested, it is also
shown that that these prosthetics will need to get both lighter and stronger in the near future in
order to more closely mimic the abilities of a real human arm. The main conclusion drawn from
this research is that a system in which a user starts with a very simple prosthetic and gradually
advances to more complex models could be highly effective in helping a user achieve a near
100% success rate in completing actions with the prosthetic, and that different combinations of
actuators (such as pneumatic and electric) would result in a more effective prosthetic.

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 2
Introduction
Todays technology is evolving at a very rapid pace when compared to breakthroughs
made under a century ago. In the 1940s, the first computer was invented with the ability to solve
simple math problems and a 22 bit-rate. In 2016, the smartphones in peoples pockets are
wireless, hand-held, and boast a 64 bit-rate, and our modernized computers can locate and collect
billions of pieces in data in the matter of a second- or less. Despite the incredible speed at which
humans can seemingly innovate technology faster than it can be consumed, prosthetics have not
seen much change since they were first used as far back as 424 B.C. by ancient civilizations like
the Romans and Persians (Norton). The biggest problem with prosthetics is that after all these
years they are still unable of doing the one thing that the parts they replace could do: move. How
effective is a replacement arm that cant do anything a real arm can? Fortunately, researchers in
the topic of prosthetics have begun to experiment with the combining of prosthetics and robotic
technology, yielding products that are capable of mimicking human movement and being
controlled neurologically, or by thoughts. A thing of the future, seen only in science fiction
movies, is actually something that will be entering the world as a reality in a very short time. The
idea of robotic body parts may still be something that seems too unrealistic or maybe even a bit
scary to some people, but this paper will seek to bring the concept of a robotic prosthetic back
down to Earth by providing insight on just how such a thing is possible, as well as what its future
holds.
Literature Review
From physical deformities and birth defects to accidents or injuries, the need for
prosthetics has been valued for as long as the technology has existed. Prosthetics allow people to
retain as much of their normal physical ability as possible in order to continue performing their

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 3
daily activities. Unfortunately, with the current technology that exists within the world of
prosthetics, amputees are given simple plastic limbs that are not capable of any movement. The
amputees ability to perform daily tasks is severely inhibited as a result. Although science and
technology has been expanding rapidly over the past few decades, the innovation of the
prosthetic is still something that is in its early stages and unavailable to the consumer market.
Many prototypes for a robotic prosthetic, which would allow a prosthetic to move on its own and
mimic a real human arm, have been tested and found successful, but there is still plenty of room
for improvement. Different types of actuators, like electric, pneumatic, and hydraulic, can be
combined to create a prosthetic arm that is capable of mimicking human motion.
Most modern prosthetics distributed to amputees who require them do not possess the
degree of motion that authentic human limbs have. While modern prosthetics are made with
more advanced materials to increase their strength and durability, their ability to function as a
normal limb is severely inhibited by the level of technology required to produce such an
augmentation. Although the idea of a moving prosthetic arm is not a new one, it still remains on
the forefront of technology with its applications in the real world still limited to the testing and
designing of such inventions.The lack of availability of such advanced prosthetics does not
staunch the overflowing need for them to amputees. An estimated 1,400 U.S. soldiers who fought
in Iraq and Afghanistan suffered from severe limb amputations (Wallace). The current state of
the technology of prosthetics means that these veterans will have to live their entire lives never
fully possessing the abilities they once had. The hardships that amputees must cope with for their
entire lives are hardships that technology can seek to resolve simply by revolutionizing the
prosthetic limb.

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 4
Actuators, a device that converts energy into motion (Tatum 2015), is thought of as the
building block of all robotic machinery, as they are responsible for creating movement in any
direction. An actuator takes a source of power, whether it be compressed air, fluids, or electricity,
and converts it into physical motion that can range from an up-down, left-right, or openingclosing motion. A common example of the use of actuators can be seen in a backhoe, which uses
pneumatic cylinders in the arm to provide it with the force to dig through dirt and lift heavy
loads. In a sense, the arm of this backhoe is like a crude robotic arm, as it has a range of motion
similar to that of a humans if one were to bend their elbow and use their hand like a scooper, in
which the pneumatic cylinder allows the arm to move freely. So taking this into consideration,
the question could be asked: can actuators be used on a smaller scale to create actual arms? This
question is one that science has not found a definite answer for, although many researchers have
come up with strong evidence to suggest it is entirely possible by creating working prototype
models.
What actuators would be able to do what, and how would you know what to use where?
First one must consider the different types of actuators that exist, which can be categorized into
three main groups based on power source: pneumatic, electric, and hydraulic. Pneumatic
actuators, which have the most practical applications in the world from robotics to
manufacturing, include End effectors [that] can range from the common [pneumatic] cylinder to
more application-specific devices such as grippers or air springs (Heney 2015). Pneumatics
work by using compressed air to create force and movement in different directions. Typically, a
pneumatic actuator can be used in small machinery because their force output is low compared to
something like a hydraulic actuator. Hydraulic actuators are commonly found in the same uses as
pneumatic actuators except the compression of the fluids that they are powered from allow for

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 5
much higher force outputs, meaning they can only be used in larger applications. Finally,
electrical actuators are powered by electrical sources which can be ideal were portable power
sources are necessary because they can operate off of batteries. The power consumption of this
however is not renewable, unlike the other two, so an electrical actuator would not be able to
function for the same period of time that a pneumatic or hydraulic could.
Although a robotic prosthetic that is capable of functioning like a human arm has yet to
see the market, many corporations and companies have created prototypes that are very close to
achieving everything a robotic prosthetic would need to be able to do. An example of this is the
DEKA Arm, a project created by the Defense Advanced Research Projects Agency, or DARPA.
The DEKA Arm is a robotic prosthetic controlled through neural brain waves that converts
thoughts into physical movement in an arm consisting of everything from the shoulder down. It
is capable of moving in the same ways a real human arm can, as and All configurations have six
preprogrammed grip patterns and four wrist movements (Resnik 2011). The physical motion in
the arm comes from small actuators and servos built into it, allowing the arm to move at the
same angles that a real arm does. Although its capabilities in terms of motion is very similar to
that of a human, The most desired improvements mentioned by our study participants and
clinicians were to decrease the device weight and to internalize or eliminate the wires and
cables (Resnik 2014). The practicality of the arm is equally as important as the capabilities for
people who would need to integrate such a product into their daily lives. While the DEKA Arm
both resembles the human arm in appearance and capabilities, researchers have also explored the
possibilities of less human arms that could accomplish the same thing as the DEKA Arm. One
group of researchers ...determined that a hand with three fingers, which have three degrees of
freedom on each finger, would be sufficient to provide stable grasping capable of force closure

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 6
on arbitrary objects (Kim 2011). For researchers, the most important aspect of a robotic
prosthetic is its functional ability as opposed to aesthetics. Ultimately, if two arms can solve the
same problem equally as efficiently (or even better than another), the appearance of the arm itself
is of little consequence. To do this, however, one must ask: what difference will make one arm
better than another? When designing these prosthetics, different actuators are applied in different
ways to create movement in certain directions. Some companies have even narrowed their focus
to certain parts of the arm in order to explore all the possibilities with that part. The problem with
this can be that different actuators not only create movement in different directions, but can also
be different sizes and have varying force outputs. These variations can result in certain aspects of
a design being weak or unproportional compared to a real human arm.
In conclusion, actuators play a major role in the design and manufacturing phase of
creating robotic prosthetics. They give the prosthetic a range of motion and allow it to move
freely in a fashion similar to how a real human arm moves. While the capability of the prosthetic
is the most important factor for how effectively it can serve an amputee, other factors, such as the
weight and power source of the arm would still require attention in order to produce the most
effective product. Engineers must first focus on maximizing the capability of the arm before
going back to optimize other aspects of it. When actuators are applied to something such as the
concept of a biomimetic arm, they can be used in such a fashion that effectively mimics human
motion so that a device that could replicate and replace a human arm could be created.
Data Collection
The topic of robotic prosthetics is filled with a plethora of new and constantly updating
academic journals and other sources about the topic as new technology is continuously created
and innovated. As a result, it is easy to find studies recently conducted by others and assess and

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 7
compare their approach to creating a solution as well as their limitations and future possibilities.
Through meta-analysis multiple academic journals and scientific research papers were collected
and analyzed to produce original research to answer the question about how actuators could be
effectively used in creating robotic prosthetics. Since the robotic prosthetic is a fairly new piece
of technology, it would not be possible to get access to it personally, so analyzing the research of
others who have experimented with this new technology was the most effective way to get
accurate and relative information about robotic prosthetics.
The question on how to design a robotic arm that can mimic the motion of a regular
human one is something that has been to an extent already been answered by science. Building a
functioning device that can recreate human motions is a challenge successfully completed by
many engineers and companies worldwide. However, the idea of connecting such a device to a
human body and having it replace a natural human limb is a topic researchers still struggle to
understand. Having the brain recognize and operate a piece of machinery as if it were a body part
made of flesh and bone is a very complicated task, and one that is answered more by devices
capable of reading brainwaves and neural impulses than it is by a robotic arm (Carmena). So
how does this relate to the correlation between actuators and robotic prosthetics? Through study
and experimentation it has been shown that the brains capability of learning how to control
actuators is fairly limited in the sense that one can only learn so much in a set period of time
(Carmena). A person would have to slowly learn simple movements and then build their way up
to more complex ones. There has also been a struggle in creating robotic prosthetics that can
maintain the size and weight of a human arm while having the same force output, as small servo
motors and actuators that would be needed to create a robotic hand can not generate the same
amount of force that human fingers and hands can (Presher). This obstacle can result in robotic

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 8
arms that are too big or heavy for the average human to be able to use comfortably. The question
of how the arm should be powered is also of concern, as a power source would have to be carried
by the operator at all times. The most viable options have worked with electric actuators so that a
battery can be incorporated into the arm, allowing for the most mobility and practicality out of
the arm (Astaras, Moustakas, Athanasiou & Gogoussis). Test subjects have been able to complete
tasks such as touching an object that requires complex hand movements or picking up a bottle
and drinking from it at a very high success rate, and has already even allowed some people to
complete tasks that they have been unable to do in their normal lives (Hochberg). By examining
the studies done on testing different prosthetic arm models, it is possible to make inferences
about how certain ideas and concepts work better than others, and how improvements can be
made on currently existing designs to make them even better and more advanced in hopes that
one day a robotic arm can work as seamlessly as a normal human arm. If I were to change
anything about the sources used to collect data from, I would try to find actual blueprints of some
of the arms used in different experiments to see if their design had anything that was reflected in
the results of the tests.
While reading and annotating the four scientific papers that were used in the metaanalysis collection process, the most important questions that looked to be answered were What
actuators are they using now? and How are these actuators effectively solving the problem at
hand, and what could be changed or improved to make them work better?. Through answering
these questions with information found in each individual source conclusions could be drawn, as
each source approached the same topic with a slightly different viewpoint on it. The first source,
which addresses the advancements made in the field of neuroprosthetics, was written by Jose M.
Carmena, a professor of electrical engineering and neuroscience at the University of Berkeley,

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 9
California. The source mainly focuses on the neural aspect of connecting a robotic prosthetic
device to the human body in such a way that the brain and recognize and interface with it,
controlling it as if it were a natural extension of the body. He states that teaching the brain how to
control the actuators in the arm is a slow process that would require time to effectively be able to
operate the prosthetic. He also discusses how simple motions that dont require much movement
or applications of different muscles are easier to learn and master than full hand and arm
movements that involve many different parts of the arm, as the brain needs time to learn how to
work each individual actuator to be able to control the movements they produce (Carmena). This
ties back to the questions being asked by creating a relationship between the amount of actuators
in the prosthetic (or the complexity of the device) and the ease in which it can be mastered by a
user, meaning it could be easier for a user to master a prosthetic if it is simpler and involves less
actuators. The second analyzed source focused on a robotic finger created by a group of
researchers known as the SMAC finger prototype. The source talked about the advantages of the
finger but also discussed the drawbacks of the finger, which included the fingers
unproportionally large size in relation to its small force output when compared to the average
human finger. This was mostly due to the direct servo technology used to actuate the finger
(Presher) being too large to make the finger smaller, suggesting that smaller actuators would
need to be created in order to create realistic fingers that can be applied to full prosthetic limbs. A
third source written by a group of researchers who all study biomedical engineering and robotics
developed and tested a prosthetic arm, comparing its motions and capabilities to that of a real
human arm. While it was capable of moving at all the same speeds and angles that a real human
arm can, it weighed much more than the average human arm, had a smaller wing-span, and ran
solely off of electric actuators. It was also only tested using very small loads to prevent damage

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 10
to the prosthetic (Astaras, A., Moustakas, N., Athanasiou, A., & Gogoussis, A), indicating that
some stronger actuators may need to be incorporated into the arm, and that electric actuators may
be capable of generating all the same movements of a real human arm but incapable of exerting
large forces. Finally, a study by Leigh Hochberg, a professor of Neurology and Engineering at
Brown University, documented and examined the data collected from testing the DEKA Arm on
two patients who suffered from tetraplegia, an illness that rendered both patients incapable of
moving their arms. The tests showed that both patients were capable of operating both the DEKA
Arm and the DLR Light-Weight Robot III after only a few training sessions and with no prior
experience at a relatively high success rate (Hochberg). However, this relatively high success
rate was only high in terms of the patients only learning how to control the arm in a few days
before testing, as the success rate for completing different tasks ranged from 44%-67%
(Hochberg). The conclusive evidence provided in here shows that even more time spent training
and learning how to control the arm could yield even higher success rates, but also raises the
question of whether using a system of sorts in which the users learn how to work a very basic
prosthetic and gradually moving up to more complex ones could have provided the same if not
better results during testing.
Data Analysis
Through the analysis and examination of research that has already been conducted on
biomimetic prosthetics, I have been able to draw from it my own information from what already
exists to answer the question of how actuators can be used to create an effective robotic
prosthetic. The most important idea to consider is that the human brain requires training to be
able to learn how to control each individual joint of the arm, such as each knuckle on a finger,
the wrist, elbow, and shoulder (Carmena). In order to get a realistic range of motion these joints

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 11
would all need actuators, meaning they will all require time to learn how to control. Thus, more
complex arms that can provide the user with more freedom and capability will take much more
time and training to use properly compared to a very simple arm design that only involves a few
motions. This has been shown in another source by the removal of wrist yaw from a prosthetic
design, as it allowed the arm to be more simple and easier to use without having a major impact
on the abilities of the arm (Astaras, Moustakas, Athanasiou & Gogoussis). Learning how to
control an arm would be very similar to learning math; no one could possibly hope to understand
calculus until they understand addition, subtraction, division, multiplication, and everything in
between. However, it has also been proven that people can easily control complex arms that
provide a full range of motion, such as the DEKA Arm, after only a single day of training
sessions involving learning how to control movement through neural signals (Hochberg), so the
effect to which this learning curve would impact someones capability in operating a robotic arm
would require more experimentation and data to understand completely. It should be noted that,
although the patients in this test could operate the arm in a very short amount of time, they could
only do so at varying success rates, whereas everyday life would require these tasks to be doable
at nearly a 100% success rate (Hochberg). Data on different robotic arms indicate that, while an
arm with a degree of motion very similar to that of a human can be made, it can only be done so
when the resulting arm weighs roughly 13 pounds (Astaras, Moustakas, Athanasiou &
Gogoussis). For a body part that typically weighs 5% of a persons total weight, this is much too
heavy to be usable for most adults. This same problem is reflected in other sources, where human
fingers that can effectively model a real finger are nearly 1.5 times the size of a regular finger,
while still not being capable of having the same force output (Presher). This is a significant
problem in trying to create robotic prosthetics, as limbs that are not proportional to ones body

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 12
will not be able to be used by a large portion of the consumer market. This is something that
would most easily be fixed through the reduction in size of the actuators and servo motors
composing the hand where possible. This is also something that would need to be done while at
the same time increasing the load that arms are capable of lifting, for as of now most tests
involve very light weights to prevent damaging the arm (Astaras, Moustakas, Athanasiou &
Gogoussis). This is a problem that would most likely be solved by using different types of
actuators that can provide stronger forces such as hydraulic or pneumatic, as opposed to the types
of prosthetics that run off electrical actuators since batteries are the most portable and userfriendly power source available (Hochberg). Current testings of robotic arms have yielded proof
that prosthetic arms can be created to complete daily tasks with high rates of success while being
easy to operate, but at the costs of practicality in size and power. The complexity of the arm has a
direct effect on how long it takes someone to become proficient at operating it, however the
degree to which this applies to operators is still not clear. However, the greatest possibility of
minimizing this learning curve while working towards gaining 100% accuracy in movements
with the arm could involve starting prosthetic users with a very simplistic arm and slowly
introducing them to more complex arms capable of more complex tasks gradually.
Conclusion
The main conclusions drawn from this research are as follows: the actuators and devices
used to actuate the movement in a prosthetic arm need to be smaller and stronger. Electric
actuators, while the most convenient in terms of power source, are not capable of creating the
forces required for daily life and need to be paired with other types of actuators (most likely
pneumatic) to give the arm added strength and ability. A graduated system of prosthetic arms
might be the most efficient for helping users learn them quickly while developing near 100%

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 13
success and accuracy rate with movements and actions. These were drawn from the facts that
most prosthetic arms designed and tested by researchers were either too short, too heavy, too
weak to lift heavy loads, or a combination of all three. Sources also made it evident that the
power source was a big issue in making a practical arm, and while battery-powered electric
actuators suited this aspect the best, they were not capable of solving every solution alone.
Finally, some sources indicated that teaching the brain how to control a robotic arm would
require a substantial amount of time before mastery could be achieved, making a possible
solution a progression-type system in which prosthetic users start with basic arms and move to
more complex ones in order to learn and maintain a high rate of success in completing more
complex tasks. The information and research discussed in this paper will be transposed to a
website in order to help it reach more people and educate them about the topic of biomimetic
prosthetics, and to help pave the road for the future in hopes that other students and teens might
take interest in the topic and pursue it as well.

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 14
References
Astaras, A., Moustakas, N., Athanasiou, A., & Gogoussis, A. (2013). Towards braincomputer interface control of a 6-degree-of-freedom robotic arm using dry EEG
electrodes. Advances in Human-Computer Interaction.
Carmena, J. M. (2013). Advances in neuroprosthetic learning and control. PLoS Biology,
11(5). doi: http://dx.doi.org/10.1371/journal.pbio.1001561
Heney, P. (12 September 2012). What is pneumatics?. Web. 10 Dec. 2015. Retrieved
from http://www.pneumatictips.com/2346/2012/09/engineering-basics/what-ispneumatics/
Hochberg, Leigh R., et al. Reach and grasp by people with tetraplegia using a neurally controlled
robotic arm. Nature 485.7398 (2012): 372+. Science In Context. Web. 4 Apr. 2016.
Kim, E-H., Lee, S-W., & Lee, Y-K. (2011). A dexterous robot hand with a bio-mimetic
mechanism. International Journal of Precision Engineering and Manufacturing, 12(2),
227-235. doi: 10.1007/s12541-011-0031-x.
Niman, J. (2013, May 2). Prosthetic technology and human enhancement: benefits, concerns,
and regulatory schemes pt.1. http://ieet.org/index.php/IEET/more/Niman20130502
Norton, K.M. (2007). A brief history of prosthetics. Amputation Coalition, Vol. 17(7), 11-13.
Presher, A. (2015, August). Novel motors give robotic hands a more human touch: new
small partial motor design offers very high power, torque and control flexibility for
implementing robotic fingers and hands. Design News, 70(8), F1+.

ACTUATORS IN BIOMIMETIC PROSTHETICS


Leonard 15
Resnik, L., Klinger, S. L., & Etter, K. (2014). The DEKA Arm: Its features, functionality, and
evolution during the Veterans Affairs Study to optimize the DEKA Arm. Prosthetics and
Orthotics International, 38(6). doi: 10.1177/0309364613506913
Resnik, L., Klinger S. L., & Etter, K. (2014). User and clinician perspectives on DEKA
Arm: Results of VA study to optimize DEKA Arm. Journal of Rehabilitation Research &
Development, 57(1), 27-38. doi: http://dx.doi.org/10.1682/JRRD.2013.03.0068.
Tatum, M. (17 November 2015). What is an actuator?. Web. 10 Dec. 2015. Retrieved
from http://www.wisegeek.org/what-is-an-actuator.htm
Wallace, D. (2012). Trends in traumatic amputation in Allied Forces in Iraq and
Afghanistan. Journal of Military and Veterans Health, Volume 20(No. 2). Retrieved
from http://jmvh.org/article/trends-in-traumatic-limb-amputation-in-allied-forces-in-iraqand-afghanistan/

You might also like