Engineers have long looked to nature for clues that will help then build robots that move with anything close to the grace that living things exhibit. Although the use of rigid metal and plastic parts tends to result in stiff, mechanical motion, a team at the Massachusetts Institute of Technology (M.I.T.) is experimenting with the use of a single piece of flexible silicon and urethane polymer to create robotic fish that smoothly wriggle through the water much like their natural counterparts.

Fish propel themselves by contracting muscles on either side of their bodies, generating a wave that travels from head to tail. To mimic the motion, the M.I.T. researchers have created two different types of robo-fish.

The first type of aquatic automaton, which measures about 12.7 centimeters, mimics the carangiform swimming technique used by bass and trout. Most of the movement takes place in the tail end of the body, says Pablo Valdivia y Alvarado, a research affiliate working in M.I.T.'s Mechatronics Research Laboratory who has teamed with Kamal Youcef-Toumi, an M.I.T. mechanical engineering professor and the lab's director. Fish that use this type of motion are generally fast swimmers, he adds.

The second type is a 20-centimeter-long robo-fish designed to move more like a tuna or shark, which swim faster and for longer distances. The motion of these fishes (and dolphins, too) is concentrated in the tail and the region where the tail attaches to the body.

M.I.T.'s robo-fish prototypes cut through the water at close to one body length per second—fast for a robotic fish, but no match for the genuine article, which can swim as fast as 10 times their body length per second, according to the researchers.

The robo-fish project is an extension of Valdivia y Alvarado's doctoral thesis work at M.I.T., which sought to create "a methodology for building mobile robots by exploiting the natural vibration of compliant bodies," he says. When people build robots, whether it is a humanoid, a quadruped or a fish, they tend to create very complex mechanisms, and this complexity often works against them by creating multiple points of potential failure, Valdivia y Alvarado says. Because the robo-fish that he and Youcef-Toumi have created is made from a single piece of polymer, it is easier to make watertight than previous generations of robotic fish.

Valdivia y Alvarado and Youcef-Toumi benefited from a wealth of data produced by previous robotic fish research. In 1994 M.I.T. ocean engineers demonstrated Robo-tuna, a 1.2-meter-long robotic fish with 2,843 parts controlled by six motors. Valdivia y Alvarado's robo-fish are powered by a single motor and are made of fewer than 10 parts, including the body and the wiring. Other robotic fish experiments have been conducted by the University of Essex in England, the California Institute of Technology and several others.

Whereas Robo-tuna and several of those that followed in its wake were built to help researchers study fluid dynamics and how fish swim, Valdivia y Alvarado hopes his new generation of robo-fish will lead to autonomous underwater devices that can perform a number of jobs, such as inspecting submerged vessels and oil and gas pipes, patrolling waterways, and detecting environmental pollutants. His initial prototypes are wired to an external power source, but the future will include battery-powered models.

And basic fish are just the first step. By December, Valdivia y Alvarado and Youcef-Toumi plan to have built a prototype manta ray and salamander, both of which require more complex movements than the original robo-fish.