A palm-sized robot — no heavier than a deck of playing cards — launches itself across grass, skips over sand, and drops into water, all without missing a beat. For something running on a modest little motor, it moves with unsettling ease.
The machine draws its inspiration from frogs, but its real secret isn’t muscle or raw power. Researchers found a way to store and release energy through bent elastic rods that snap — a geometric trick that does the heavy lifting so the motor barely has to.
A frog’s leap, distilled into engineering
Frogs aren’t the first animals to inspire robot designers. Horses, insects, and countless other creatures have shaped how engineers think about movement and propulsion. But frogs offered something specific: a model for storing energy and releasing it in a single, explosive burst.
Researchers at the University of Michigan and UCLA studied exactly that mechanism, focusing on what happens when you bend an elastic rod and then twist its ends. Published September 18 in Science Advances, the study explores how that combination produces a powerful snapping motion — and how a small robot can harness it repeatedly.
The key insight, as co-first author Xiaonan Huang put it, is letting the robot’s own mechanics do the work. Rather than demanding constant high output from a motor, the goal is to program an elastic structure to store energy and release it rapidly — giving a small robot access to powerful, repeatable motion on the cheap.
The physics of a snap
Not every combination of bending and twisting produces a snap. Some configurations cause a rod to change shape gradually, with no sudden release. Others force it to flip abruptly from one stable shape to another — and that abrupt transition is where the useful energy lives.
To find the right combination, the team ran computer models alongside hands-on experiments with a robotic arm, repeatedly deforming rods in different configurations and looking for the geometry that maximized energy release. What they landed on was a helix — a coiled-spring shape — that produced the strongest, most reliable snap. It turns a relatively small motor rotation into a sudden, forceful burst. That’s the mechanical trick at the heart of the whole design.
Building the robot: from simulation to prototype
The prototype is small enough to sit in your palm and weighs just 3.4 ounces — roughly the same as a deck of playing cards. Despite that modest size, it moves at about three body lengths per second, launching itself forward in short, rapid hops.
Two snapping rods are mounted at the rear of the robot’s body, each connected to a rotating motor. The motor winds the twist into the rod until it snaps, then unwinds and reloads, producing a continuous hopping cycle without requiring the motor to work particularly hard. Steering is built into the same system: activating only one rod at a time causes the robot to turn, which was enough for the team to demonstrate remote-controlled navigation through a small obstacle course.
Tested on wood, sand, grass — and water
The researchers didn’t limit their testing to a lab bench. The robot was put through its paces on wood, glass, leather, sand, and grass — surfaces that each present different friction, compliance, and unpredictability. It handled them all.
Outdoor trials pushed further. The robot climbed and descended steps, a meaningful hurdle for any small hopping machine. Then, with paddle attachments added, it demonstrated the ability to swim. That combination — reliable hopping on land and functional swimming in water — is unusual at this scale. Most small robots are optimized for one environment. This one moves between them without a fundamental redesign.
What this means for the future of small robots
One of the more significant aspects of this work is how well the design principles scale. Because the snapping behavior depends on the rod’s shape and how its ends are moved, the same rules apply across different sizes. Co-first author Khalid Jawed noted this opens a path toward robots just a few millimeters in size — machines that could navigate spaces far too small for anything currently available.
The efficiency insight matters just as much. Offloading mechanical work to the robot’s structure rather than its motor means smaller, lighter, lower-power systems can still produce meaningful force — which has real consequences for deployment outside a lab. Researchers point to search-and-rescue operations, environmental monitoring, and exploration of hazardous or hard-to-reach spaces as potential applications. Whether the design eventually reaches those contexts depends on further development, but the underlying principle is now demonstrated and published. The next step is scaling it down, and seeing how small this idea can actually go.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
