Strength in Innovation

In the 1970s, the Bionic Man was merely a figment of the imagination, dreamt up by television writers. Today, engineers and researchers are finding ways to construct the very essence of the human body with artificial muscles.

Gripping technology

The term artificial muscle refers to any device that expands and contracts with stimulation, performing functions similar to those of a biological muscle. Different types of artificial muscle technology already exist and are being used in some form or another.

One of the promising types of this technology is dielectric elastomer, also known as electroactive polymer artificial muscle (EPAM). According to Roy Kornbluh, senior research engineer for the engineering and systems division of SRI International in Menlo Park, Calif., this electroactive polymer (EAP), which was invented at SRI, works based on an electrostatic effect called Maxwell Stress, where a conductive material is applied to a thin, rubbery film. When voltage is applied to that film, with a positive charge on one side and a negative charge on the other, the charges are attracted to each other and squeeze the film. When the film contracts in thickness, it expands in area, creating an action similar to muscle flexing.

“Pull your muscle [to] show how strong you are and your muscle [grows], but actually [it] contracts in length and becomes large in diameter,” said Yoseph Bar-Cohen, PhD, senior research scientist and supervisor of the Advanced Technologies Group at NASA’s Jet Propulsion Laboratory in Pasadena, Calif. “It’s the same kind of mechanism in an insect and in an elephant.”

Image reprinted with permission of Yoseph Bar-Cohen, PhD, Jet Propulsion Laboratory/NASA/California Institute of Technology

Still a fairly young technology, this innovation has been used to produce devices that are relatively small, but researchers are investigating the possibility of creating far larger objects.

“It would be hard to make something at the size scale of a bicep at this point, but certainly that is something we are working toward in the future,” Kornbluh said. “There are not any fundamental limitations why we cannot get to the force levels needed for even a strong muscle like a bicep, or at some point in the future, maybe even some muscles for the leg.”

The EPAM technology functions with a high voltage but, because it is maintained at a relatively low current with thin wiring, Kornbluh expects a number of protective measures could be built into the device to ensure the user’s safety. In addition, the technology is electrically efficient, which would conserve battery supply. Pneumatic muscles, another type of artificial muscle technology, are tubes that operate with the force of pressurized air. Because of a cloth-like sleeve surrounding the tube, airflow is restricted in expansion. As the tube expands, it shrinks in length.

Synthetic muscles powered by hydraulics have characteristics similar to pneumatic muscles, but instead are powered by fluids. Both pneumatic and hydraulic artificial muscles are slower, louder and more difficult to control than electroactive polymer muscles. For these reasons, pneumatic and hydraulic muscles are useful primarily for heavy machinery and equipment.

Piezoceramics is another type of actuator where electrical energy is applied to a ceramic crystal, whereby the crystal changes shape. This technology uses high voltages to create a small amount of movement.

“You could lift your car one-5,000th of an inch, [but] that is different than moving an elbow through 60° or 70° in a rapid time and lifting a weight,” said Bradley Veatch, PE, senior research engineer for ADA Technologies Inc. in Littleton, Colo. He also noted the use of nichrome wire as an actuator.

“Its difficulty [is that] it works well in terms of operating scale, but it is relatively slow, not energy efficient and it generates a great deal of heat. That can be a problem because you tend to melt things,” he said. “We are trying to overcome impedance mismatch and at the same time, address thermal and efficiency issues.”

Imitating life

With a background in ultrasound, Bar-Cohen became a pioneer in artificial muscle development by accident after finding an article that was full of incorrect information about such material.

“At that time, there was no community of experts that you could take a paper like that and show it to them and they could say there is no way it would work. There were people who reviewed it, [but no experts] to realize that the paper was wrong.”

Bar-Cohen set out to establish such a community. In 1999, he founded the Electroactive Polymer Actuators and Devices conference through the technical society The International Society for Optical Engineering, as a place for researchers and engineers to discuss findings in the various fields where artificial muscles play a part. The community has since grown to include more than 1,300 people around the world.

Kornbluh recognized the need for artificial muscle technology in robotics in the early stages of his research. The technology was originally envisioned as a way to help create new generations of robots that could be more like humans, including natural muscles.

“Coming from the mechanical side of things, the problem is that you are trying to make something imitate muscles with motors and gears,” Kornbluh said. “How are you going to imitate muscle with motors and gears? What you want is an artificial muscle.

“If you are thinking robots, then you are also thinking prosthetics and orthotics because it is just a question of where you draw the line,” he said.

The O&P community has much to gain from advancements in this area. The technology will not only influence the manufacture of future prostheses, but also the development of replacement limbs and muscles.

“The area I think you will probably see these actuators applied to first is going to be lower extremities just because of the size of the market,” Veatch said. “I would not be surprised to see these actuator systems used for driving a knee or an ankle. We are not to a point yet where we can raise an arm or swing a knee well. We are getting there, but I don’t think we are there quite yet.”

A test system lifts a 4-pound weight against gravity from rest (far left) to almost 90° of elbow flexion (far right). The black cylinder objects in the upper part of the arm are EPAM actuators arranged in a hexagonal pattern of seven discrete units. The white cords and pulleys transform the short stroke of the EPAM actuators into the movement necessary to raise the arm. The EPAM actuators are energized to extend and retract in unison pulling on the cable/cord. Images reprinted with permission on ADA Technologies Inc.

Racing to the finish

Despite the myriad developments in the past few years, this technology is not yet fully mature.

In 2002, the first artificial muscle product was commercialized by EAMEX, Japan, in the form of a fish robot. It swims without batteries or a motor and it uses artificial muscles that simply bend upon stimulation and propel it forward in a fish tank. For power, it uses inductive coils that are energized from the top and bottom of the tank and it is charged similar to an electric toothbrush. Another product is close to the commercial market finish line: a small auto-focus mechanism, created by Artificial Muscle Inc., which will be mounted in a cell phone camera. The muscle would provide the camera with a great degree of tolerance to impact and it can survive a crash.

“You can throw it on the floor and [it] will survive that,” Bar-Cohen said. “The lens may break, but the muscle will survive it.”

Another area of interest is creating heel-strike generators operated with artificial muscle technology. Initially sponsored by the Defense Advanced Research Projects Agency (DARPA) of the U.S. government for lightweight power generators that would offer a new energy supply to soldiers, the technology eventually may be applied to shoe generators that could power cell phones or MP3 players.

“It has always been of interest for prosthetic, orthotic or other biomedical devices, since obviously you do not want those batteries to run out,” Kornbluh said. “For example, if you had a lower limb device, maybe a semi-active one that has some electric control, it would be relatively straightforward to put in a heel-strike generator because you are connected to it in the heel anyway.”

He also said his team is researching the possibility of using the technology for implantable prostheses, artificial hearts or devices for people with paralyzed breathing muscles.

“There is only so much that can be done practically at this stage of the technology because of the weakness we still have in the materials that are available to us, but there is an enormous range of what would be nice to have, down the road — [perhaps ultimately] is implants for making a bionic person,” Bar-Cohen said. “The other could be an exoskeleton, which is something you wear possibly on your leg or hand that helps you. Or it could be … active prosthetics.”

It only may be a matter of time before artificial muscles break into the world of environmental sustainability as well, with implications for conserving energy.

“If you stretch and contract [artificial muscles], you can generate electricity if you put on the voltage and take it off at the right times. We are looking at that to harvest energy from ocean waves,” Kornbluh said. “So it is kind of the opposite of a muscle, but that is allowing us to scale up the processes and make things that are going to be even bigger than muscles.”

Stretching out

The future possibilities of artificial muscle technology are endless, but researchers must be realistic with their goals.

“Development of new technology will always be difficult if it tries to replace existing technology. If we can learn from nature, if we can be inspired by nature, then we can do better than nature, as we have shown with the development of airplanes,” Bar-Cohen said. “My hope is to see more and more niche applications of artificial muscle technology.”

Veatch said the best solution is for the various professions to take a building block approach to the technology.

“I think what is needed for the field is to take a viable technology and produce it in a modular form,” Veatch said. “I think of an artificial muscle as being more of a modular unit that can be put onto a prosthesis … that functions more as a muscle. As part of the prosthesis, not as a motor that’s screwed or bolted on.”

At ADA Technologies, engineers in a spinoff group called PhysioNetics have been developing technologies that will ultimately allow for more affordable orthotic and prosthetic devices.

“We are looking at technologies that are suitable for mass distribution that work within our third-party payer system,” Veatch said. “We have developed terminal devices, [which is] our emphasis, although we also have done some work with orthotic bracing for ankle dorsiflexion problems.”

Man vs. Machine In 1999, Yoseph Bar-Cohen, PhD, senior research scientist and group supervisor of Advanced Technologies Group at NASA’s Jet Propulsion Laboratory in Pasadena, Calif., organized the first Electroactive Polymer Actuators and Devices (EAPAD) conference, establishing and bringing together the artificial muscle community through the annual Smart Structures/NDE conference sponsored by The International Society for Optical Engineering (SPIE). Bar-Cohen posed a challenge to this newly formed community: create a robot that can arm-wrestle with a human and win. “Initially the challenge was to wrestle and win against any human, but eventually it will have to be the strongest human on earth,” Bar-Cohen told O&P Business News. “At the time it sounded like a crazy idea but I was a little bit inspired by the chess game for computers,” he said. “In the 1970s, we thought there is no way a computer could win in a chess game where there is a need for mental capability that is so advanced. Now we know that a computer … can do it.” He wanted to see if the community could develop the technology to the level needed to make a person lose against a machine driven by this plastic. On March 7, 2005, three contestants emerged at SPIE’s 2005 EAPAD conference, ready to accept the challenge. Bar-Cohen chose Panna Felsen, a high school student from San Diego with an interest in robotics, to go up against the machines created by three different teams. Felsen, who founded an engineering club at her school and planned to become an engineer after graduation, was a natural choice. Felsen’s first opponent was Environmental Robots Incorporated of Albuquerque, N.M., headed up by Mohsen Shahinpoor. The robot, made of ionic polymer metal composites, put up a fight, but Felsen pinned the artificial arm within 26 seconds. Next, a robotic arm propelled by dielectric elastomers created at the Swiss Federal Laboratories for Materials Testing and Research of Dubendorf, Switzerland. To avoid electric shock, the Swiss engineers provided Felsen with a heavy rubber glove. She beat this opponent in 4 seconds. The final contestant was a team of undergraduate students from Virginia Polytechnic Institute and State University — Steven Deso, Stephen Ros and Noah Papas — along with their adviser John Cotton, an associate professor. The students created their robot as part of a class project, and built most of it from scratch because of a lack of funding. This team provided Felsen with safety goggles, but she won against the robot in 3 seconds. “Even though none of the arms won, this has been an important milestone for the field,” Bar-Cohen said. Since the current capabilities are unable to win the match, the contest was modified in 2006. A test machine was created by Bar-Cohen and members of his group, jointly with Qibing Pei, a professor at the University of California, Los Angeles, and his students to gauge the technology’s progress as advances are made. This machine, designed to support wrestling action while measuring the speed and force, was used in the second contest, held on Feb. 27, 2006, in San Diego during the EAPAD Conference. Felsen arm-wrestled with the test machine and her capability was measured to provide a reference for contesting arms. Environmental Robots Incorporated and two groups of students from Virginia Polytechnic Institute and State University participated in this contest. The three arms wrestled with the machine and the data was compared with Felsen’s speed and force. Again, she had superior capability but this time, the numbers quantified that she was 100 times faster and stronger. “It seems that this data deterred potential scientists and engineers from taking on the challenge in 2007,” Bar-Cohen said. Because of Bar-Cohen’s interest in encouraging advances in plastic actuator development, he raised the bar on materials to be used in future contests. “Eventually I would like to see competing ams with performance and characteristics much like the human arm,” Bar-Cohen said. With input and consensus from members of the 2008 SPIE EAPAD Organization Committee, he announced that beginning next year, there will be stricter rules for the artificial muscle materials that can be used to drive contesting arms. Because using electric power to drive plastic actuators is more practical than other methods, the key restriction is that no actuator that requires gas or liquid (i.e., no pneumatic or hydraulic) or adding chemicals will be allowed to participate. “The arm-wrestling match is intended to provide the public a sense for the state of the artificial muscle field,” Bar-Cohen said. “As progress is made, we should be able to handle harder challenges. Eventually, we expect to benefit greatly from emergence of effective plastic actuators. It would allow for superior prostheses, as well as human-like robots that look and operate similarly to us.” For more information: http://ndeaa.jpl.nasa.gov/nasa-nde/lommas/eap/EAP-armwrestling.htm

Breaking through barriers

The technology is developing at a steady pace, but, as with any new developments, some issues hold it back. Procuring research and development funding is the most difficult aspect of artificial muscle research.

Veatch applied for a grant to continue the research and was declined because the reviewers correctly assessed that the technology was not yet sufficiently mature.

“This struck me as odd because I was trying to get funding to mature the technology,” he said. “The biggest thing that would help is to have more funding for the basic research.”

Kornbluh has had similar experiences acquiring financial support.

“Aside from an occasional good opportunity like the DARPA program, there is not a lot of funding available from the O&P community. There is just not that kind of money around.”

Kornbluh said it would require a partnership between people who make the technology and an O&P manufacturer willing to turn it into a prosthesis.

“If that partnership materialized … I think I would say [we could make commercial] products in 5 years or so,” he said. It does not require any breakthroughs in performance, it just requires good solid engineering and making the technology more reliable. I am hoping that people read this and say, ‘Yes, we want to be part of that partnership.’”

Veatch said another issue delaying the advanced technology’s movement into the O&P profession is the advanced technology itself.

“In the O&P field, in my opinion, there tends to be a kind of reluctance to move in with advanced solutions, mainly because I don’t think that clinicians are necessarily familiar with all of them,” Veatch said. “They require training to implement, and on top of that they are costly, and there are always limitations in terms of durable medical equipment reimbursement.”

Lack of interest is not a problem for the artificial muscle community, Bar-Cohen insisted. He believes that the community needs the time to build a solid foundation.

“We need an infrastructure like every infrastructure of various other fields. We need good science. Then we need good engineering. Once that is established, then everything else is resolved.”

Bar-Cohen said the community is already making advances toward developing strong materials. Kornbluh added that the next step is to prepare the technology for the various applications.

“We have shown the performance in the laboratory. It is a question of packaging that into useful and, perhaps most importantly, reliable and safe devices. I think those are the challenges and [we need] incremental improvements. We will get there, but it is going to take some time,” Kornbluh said.

Veatch does not foresee one solution for every potential artificial muscle use.

“I do not believe that, even if we had an ideal artificial muscle, it would be the correct solution for everyone,” he said. “I think that the real future lies in hybrid technologies, where we produce building block pieces, where you could produce prostheses that are a combination of powered elements or passive elements that best suit the need. The question is, what does the person want to do?”

Bar-Cohen said one time he had 10-year-old students visit his lab. At the end of the visit, a student asked Bar-Cohen if he was a robot.

“Just the fact that he asked this question shows it is becoming a possibility and a reality,” he said. “When I was 10 years old, I didn’t even imagine that a person could be a robot, so this is a direction we probably could go into. Who knows? It could be our future companion or our future appliance.”

For more information:

• http://ndeaa.jpl.nasa.gov
• http://SPIE.org

Stephanie Z. Pavlou is a staff writer for O&P Business News.