We Can Rebuild You

The room looks like a set for one of those grade-B horror movies Ed Wood loved to make. In one corner, a skeleton draped in wires spasmodically jerks its arm up and down-meet “Mr. Bony.” In another corner, strange, gelatinous creatures undulate like jellyfish inside tanks. More of the slippery blobs wiggle inside Tupperware containers that overrun the cabinets. “On campus, they call this the spooky lab,’” notes Mohsen Shahinpoor, director of the University of New Mexico’s Intelligent Materials Laboratory. “And to tell you the truth, it can get a little scary around here.”

Shahinpoor, the genial assistant dean of the College of Engineering, is an unlikely successor to Dr. Frankenstein. Although he is not trying to create life in his laboratory, the inanimate materials he melds together squirm and writhe like living entities. His true objective is to develop a host of supple artificial muscles that may eventually run machines and robots, and perhaps replace worn-out or defective human parts.

This tantalizing prospect stems from a discovery made by Israeli scientists in the late 1940s: certain plastic-liquid mixtures called polymer gels can flex and relax like natural muscles. The slimy fiber bundles shrink when the solution they’re immersed in becomes acidic; but stir in some base and they swell to many times their former size.

The same effect can be induced by attaching electrodes to the material and passing a current through it. This works because of electrical attraction and repulsion: when polymer gels are in an acid solution, negative ions from the gel are attracted to positive ions from the acid that permeate the gel, causing the material to contract. The opposite phenomenon occurs in an alkaline solution: the material expands when its negative ions are repulsed by negative ions that have infiltrated from the solution.

Some 50 years after this conceptual breakthrough, research on artificial muscles has recently picked up thanks to progress in materials science. “We now have theories that tell us in advance how to design materials with the properties we want,” explains Daniel Segalman, a chemical engineer at Sandia Labs who collaborates with Shahinpoor. “This eliminates much of the trial and error.”

Working with artificial silk fibers and a blend of polyvinyl alcohol, polyacrylic acid, and other compounds, Shahinpoor and his Sandia colleagues have fabricated simple devices that demonstrate possible uses for synthetic muscles. For instance, they have designed small motors that harness the expansion and contraction of a polymer gel to open and close a valve, compress a spring, or rotate a pulley.

Shahinpoor (a.k.a. “Mo the Muscle Man”) is already exploring more ambitious ideas. The U.S. Navy, for instance, is interested in his swimming “robofish,” which moves by flapping its tail without relying on any mechanical parts. The Navy’s motivation is to devise a noiseless propulsion mechanism for submarines.
The Army, meanwhile, recently gave Shahinpoor a three-year contract to develop exoskeletal systems that could enhance human muscles, providing extra strength for tired soldiers. As part of that just-completed work, Shahinpoor and his colleagues attached some muscles to a jacket sleeve to supply added flexing power. A potential drawback of exoskeletal clothing or suits, according to Sandia engineer Walter Witkowski, is that “the weight can offset the extra strength it might render.”

That problem, however, could be avoided in the weightless environment of space. In fact, Shahinpoor suggests putting muscles in space suits to give astronauts more strength and dexterity in their legs, arms, and hands, and NASA is supporting his efforts. “In current space suits, you can’t close your hands easily, or bend your elbows or knees, because the pressure in the suit makes it so difficult,” he says. “We’re hoping that people will be able to move more freely with augmented power.”

Shahinpoor would also like to overhaul robots by exchanging clunky gears, motors, and pulleys for flexible, electronically controlled muscles. Mr. Bony, who can move his arm with a set of artificial biceps and triceps, is an admittedly primitive step toward the development of humanlike “bionic” robots.

Team members are also exploring several medical applications. They’re working, for example, with medical researchers at Columbia University and the University of New Mexico to study the idea of wrapping diseased hearts with artificial muscles that would keep them beating regularly in lieu of a transplant operation. The University of New Mexico has already filed patent applications for the idea, which will first be tested on animals, possibly rabbits. But it is still too early to apply for approval from the Food and Drug Administration, Shahinpoor says. Meanwhile, companies are considering the use of polymers as artificial sphincters to treat incontinence. The possibility of using artificial muscles for muscle repair or perhaps even “bionic” limbs lies further down the road.

David Brock, who heads the Artificial Muscle Project in MIT’s Artificial Intelligence Laboratory, is also pursuing several applications for polymer gels. Indeed, his lab, like its New Mexico counterpart, is inhabited by a host of muscle-driven machines-including an imitation fish that resembles a floating wig and a movable joint reminiscent of Mr. Bony’s arm. Brock cautions that many advances will be needed before any but the simplest applications become practical. Implants are especially problematic, he says, as one has to worry about biocompatibility, toxicity, and rejection when placing foreign materials inside the human body. The good news, however, is that many artificial muscles may already be composed of biocompatible materials, Brock notes, and, if not, they could likely be encapsulated in more benign substances that do not trigger an immune response.

For now, Brock is focusing on more immediate concerns, explaining that “once we come up with great polymer gels and great ways to activate them, the applications will come easily.” Brock and Woojin Lee, an engineer formerly working in his lab, have recently demonstrated a mechanical control system that can precisely regulate the motion of an artificial muscle, as well as the force it exerts. The polymer gels swell
and contract in response to chemical stimuli the same way Shahinpoor’s materials do, but the artificial muscle has been incorporated into a mechanical apparatus attached to a robotic “arm” and equipped with force and position sensors that allow an unprecedented degree of motion control. “You can tell it where to go, and it goes there,” Brock says.

Of course, if we ultimately use such artificial muscles and devices to power a new generation of robots and to reengineer humans, we will then have something new to think about: As robots become more like us, will we become more like them?