This robot is better at jumping than any machine or living thing

Frogs do it. Wallabies do it. Even educated fleas do it.

Now, a robot can out-jump them all.

In a paper published Wednesday in the peer-reviewed journal Nature, researchers present a relatively simple way to quantify and compare "jumpers" — a category that includes both living things and engineered machines — across all scales. But they didn't stop there. 

"Following these insights, we created a device that can jump over 30 [meters] high, to our knowledge far higher than previous[ly] engineered jumpers and over an order of magnitude higher than the best biological jumpers," they write. Their jumper, which stands just less than one foot (30 cm) high and weighs just more than one ounce (30 g), can launch itself 108 feet (33 m) into the air, with a take-off velocity of 92 feet (28 m) per second.

Watch it in action here:

A new model for jumping makes better jumpers possible 

Humans have long been fascinated with jumping, which the researchers define as "movement created by forces applied to the ground by the jumper, while maintaining a constant mass." (That excludes machines like rockets and arrows shot from a bow.) Aristotle discussed using weights to jump higher, and Rennaisance scholars developed a rudimentary model to analyze jumping in the animal kingdom. For more than half-a-century, engineers have looked to the biological world for inspiration in designing jumping machines. 

Jumping forces both engineers and evolution to confront some basic physical limitations of power generation. "Muscles and motors cannot generate the high power outputs necessary to propel... jumpers on their own," mechanical engineer Sarah Bergbreiter writes in a Perspective published alongside the paper in Nature. Living and engineered systems alike get around this limit by using their "muscles and motors to store energy in spring-like structures" before releasing the energy all once in a method called latch-mediated spring actuation, she writes.

The insights allowed for revolutionary designs

Earlier researchers had investigated how some of nature's most prolific jumpers (such as the tiny froghopper, an insect that can jump 115 times its body length) manage to propel themselves so high into the air, but those studies were limited by some assumptions the new study confronted head-on. For instance, there are some big differences between linear motors in living things (i.e. muscles) and the motors available to engineers.

"The rotary motors that are commonly found in engineered robots can overcome [these limitations] by acting as a winch to provide a force over a much larger displacement without needing a larger motor, thereby increasing work density," Bergbreiter says. "As long as rotary motors can continue rotating, engineered systems are instead limited by the energy density (stored energy per mass) of the springs."

The researchers also realized that a mix of rubber bands and carbon-fiber springs could store a huge amount of energy per mass. Unlike most springs, "[t]his configuration also results in the spring requiring a relatively constant compressive force to be applied over a range of distances," Bergbreiter says. That's a big advantage if you're trying to build a record-setting jumping robot. Finally, the modeling work helped the researchers see that the typical proportions of a jumper aren't necessarily optimal. Their invention features springs that are much larger than the motor.

"This unexpectedly high ratio between spring and motor mass is a result of
the relative energy limitations of these components: the energy density of the spring limits the robot’s jump height, whereas the work density of muscle limits the amount of energy that biological jumpers can store in their springs," Bergbreiter explains.