Robots That Swim and Jump on Living Muscle
A robot the size of a stick of chewing gum has swum through water on the power of living muscle.
Researchers at the Massachusetts Institute of Technology built a paper-thin swimming robot driven by a single layer of living muscle cells, they said. The results appeared Sept. 28 in the journal Advanced Functional Materials.
The robot’s skeleton is a gel film the length and width of a stick of gum and 0.5 millimeters thick. Each half of the film acts as a fin, and each fin is covered with a layer of muscle cells far thinner than a human hair. One fin measures 15 millimeters long and 7 millimeters wide.
The muscle cells were genetically engineered to contract in response to light. Shining light on one fin moves the muscle on that side and flutters the fin, and alternating light between the two sides steers speed and direction. With no circuits or wiring, the robot is controlled entirely by light from outside.
The team filled a large culture dish with water, set up a maze and steered the robot by moving light by hand. Its top speed was about four body lengths per minute, slow next to an Olympic swimmer’s 65 body lengths per minute but similar to a leisurely cow shark, the researchers said.
“Moving through water takes much more force than moving through air,” said Ritu Raman, associate professor of mechanical engineering at MIT, who led the research. “The robot is quite strong for its size.” The team said it is the first case of a self-moving, thin, two-dimensional muscle robot.
Older biohybrid robots attached whole lab-grown muscle masses, requiring millions of cells. Growing muscle as a thin layer cuts production costs and improves the efficiency of movement. The team grew muscle cells aligned along square grooves on a stiff gelatin-based gel and strengthened them by repeatedly exposing them to light. According to the paper, the thin muscle produced about 20 times the force per unit volume of conventional three-dimensional muscle, and the robot operated for more than 30 days.
Living tissue is soft, responsive to its surroundings and able to heal itself, the researchers said, making it suited to delicate work in fragile environments where conventional machines cannot go.
“The next goal is to optimize the body design so it swims faster,” Raman said. “Even at slow speeds, you can picture a muscle-driven swimming robot being used for underwater environmental monitoring.”
The research was partly funded by the U.S. Office of Naval Research. The paper’s first author is doctoral researcher Mahira Bawa.
Less than a month after the MIT announcement, similar news came from Switzerland. Researchers at ETH Zurich unveiled small robots powered by living skeletal muscle on Oct. 22, with results published in the journal Nature Communications.
The Zurich robots jump, swim, walk and grasp. They combine muscle tissue grown from rat cells with a water-rich hydrogel scaffold, and a single actuator is about 1.5 centimeters long. One contraction lifts a robot about 5 millimeters, and a walking robot moved 6.2 millimeters per second.
Both studies arrived at the same answer. Lab-grown muscle often contracts too weakly to produce visible movement, but the bottleneck was not the amount of muscle. What mattered was how efficiently the scaffold converted the muscle’s force into motion.
Where MIT solved the problem with groove shapes and muscle training, the Zurich team designed muscle and scaffold together from the start in computer simulations, gaining up to 11 times the range of motion from the same amount of muscle. “Instead of fixing the shape of the muscle in advance, we wanted to design the whole thing from scratch on a computer,” said co-first author Aiste Valciunaite, a doctoral researcher in the Soft Robotics Lab led by Professor Robert Katzschmann.
Biohybrid robots are still confined to the laboratory. They are slow, their muscles tire, and they must be supplied with nutrients. MIT’s robot manages four body lengths a minute and depends on external light for steering.
What the two studies together show is that the cost structure of the field may change. Growing muscle cells takes time and money, and until now a usable movement required a muscle mass made of millions of cells. If a single thin layer of muscle, or a small amount of it, can produce useful movement, production costs fall sharply.
Living tissue also has qualities no conventional machine has: softness, responsiveness and self-repair. That is part of why the Office of Naval Research funded the MIT work. A quiet, soft underwater explorer could work where metal and motors cannot go.
Underwater environmental monitoring, as Raman pictures it, is the nearest application. A product is years away, but the method of designing muscle as a machine component took a step forward this month in both studies.
