The Dual Medium Nightmare
Air and water do not mix. For engineers building flying machines, water is the ultimate enemy. It is roughly 1,000 times denser than air. If you try to spin a high-speed drone propeller underwater, the motor will burn out in seconds. If you build a heavy submarine, it will never leave the ground. The physics of the two environments are so fundamentally opposed that we have spent a century building entirely separate machines for them.
Then you look at a puffin.
These odd little seabirds fly through the sky at highway speeds, plunge directly into the ocean to hunt fish, and then burst right back into the air to fly home. They use the exact same pair of wings for both tasks.
For years, roboticists assumed replicating this feat required complex, heavy engineering. We thought we needed folding wings, multiple motors, and complex shifting gears. We were wrong.
A team of researchers from MIT and EPFL in Switzerland just proved that the secret to masterfully navigating both air and water is actually extreme simplicity. They built a 250-gram robot called the FAAV (flapping-wing aerial-aquatic vehicle) that swims underwater, breaks the surface, and flies through the sky.
It does all of this with just two wings and a single motor. No propellers. No folding joints. No paddling feet.
This little machine represents a massive shift in how we design dual-medium vehicles. It shows that instead of fighting the massive difference in density between air and water with heavy hardware, you can let the physics of materials do the work for you.
The Secret Is Letting the Wings Bend
When a bird dives into the water, it folds its wings close to its body to reduce drag and protect its bones. Early attempts at dual-medium robots tried to copy this folding mechanism. But adding hinges, joints, and extra motors to a tiny robot makes it heavy. Weight is the absolute killer of flight.
The MIT and EPFL team, led by Raphael Zufferey and Moritz Hüsser, found a brilliant workaround. They skipped the folding joints entirely. Instead, they designed wings with passive flexibility.
The wings are made of a very thin, translucent nylon membrane reinforced with carbon fiber struts. The surface is coated with hydrophobic nanoparticles to shed water instantly.
When the robot is in the air, these wings are stiff enough to catch the wind and generate lift. But the moment the robot enters the water, the extreme density of the liquid pushes against the wings, causing them to bend passively by up to 90%.
This bending is the key. By deforming so drastically underwater, the wings effectively shrink their stroke. The motor doesn't have to fight the heavy water. The wings simply slide through the liquid with minimal resistance, saving the battery and protecting the internal gears.
It is an incredibly elegant design. The robot does not need to "think" about changing its wing shape. The physical properties of the water force the wings to adapt automatically.
Escaping the Grip of Surface Tension
The hardest part of the entire journey is not swimming or flying. It is the transition between the two.
When a robot tries to break out of the water, it faces a brutal physical challenge. It must accelerate upward while its body is still partially submerged, shedding heavy water while its wings struggle to find purchase in the thin air above. If the wings hit the water's surface at the wrong angle, the water grabs them and drags the robot back down.
During testing in a water tank and eventually in the open waters of Lake Geneva, the team discovered that the launch requires highly precise geometry.
The robot must approach the surface at a steep pitch angle of exactly 70 degrees.
/ <- 70-degree launch angle (prevents wingtips from dragging)
/
/____ Water Surface
If the angle is too shallow, the trailing edge of the wings or the tail will drag in the water, creating too much drag and causing the robot to stall. If the angle is any steeper than 70 degrees, the robot simply flips backward and crashes back into the lake.
When the robot hits that sweet spot, the entire transition takes less than one second. The body breaks the surface, the wings complete eight to ten rapid strokes, and the robot is suddenly airborne, cruising at about 6 meters per second.
Biologists Are Shaking Their Heads
This robot is doing more than just showing off engineering talent. It is actually teaching biologists how real animals work.
When you watch a puffin, a duck, or a loon take off from the water, they almost always paddle furiously with their feet. They run across the surface of the water, flapping their wings and kicking their legs to build up enough speed to break free of the water's surface tension.
Biologists assumed this leg-paddling was an absolute requirement for birds of that size to take off from water.
The FAAV robot proved them wrong.
The robot has no legs. It cannot paddle. Yet, using only its medium-sized, flexible wings and its motorized tail, it successfully launches itself directly out of the water and into the air.
This reveals that puffins and other diving birds might not actually need to use their feet to escape the water. The feet certainly help, but the physics of flapping wings alone is completely sufficient if you get the wing flexibility and the launch angle right.
Using a highly controllable robot allows researchers to test wild theories that you could never test on live animals. You can't program a live puffin to flap its wings at a bizarrely slow frequency or force it to try taking off without using its feet. But you can do exactly that with a robot, revealing the absolute limits of animal biomechanics.
How to Avoid the Heavy Shell Trap
If you want to build a small robot that flies, every single gram matters. The FAAV weighs just 250 grams. That is roughly the weight of a single cup of coffee.
Normally, waterproofing a robot means putting all the sensitive electronics inside a hard, sealed plastic or metal box. But those boxes are heavy. If the team had used a traditional waterproof housing, the robot would have been far too heavy to lift itself into the air.
To solve this, the engineers coated each internal component individually with flexible silicone.
The battery, the tiny electric motor, and the custom crankshaft are all sealed in their own protective layers. This keeps the robot incredibly light. It also makes the robot neutrally buoyant.
Neutral buoyancy means the robot does not float to the top like a cork, nor does it sink to the bottom like a stone. It simply hovers in place underwater. This is a massive energy saver. The tiny battery doesn't have to waste energy fighting gravity or buoyancy while the robot is submerged. It can focus all of its limited power on moving forward.
What This Means for the Future of Ocean Research
Right now, collecting data from the ocean is a slow, incredibly expensive process.
If scientists want to measure water temperature near an iceberg, check on a coral reef, or track an algae bloom, they usually have to charter a boat. That requires a crew, fuel, safety gear, and hours of slow travel. It costs thousands of dollars a day.
Standard aerial drones don't help much because they can't get wet. Submersibles don't help because they take forever to travel long distances.
A bird-like robot changes the math entirely.
On a single battery charge, the FAAV can fly about 6 kilometers (3.7 miles) or swim about 2 kilometers (1.2 miles). Because flying is far more energy-efficient than swimming through dense water, the most efficient way for the robot to operate is to fly to its destination, plunge into the water, take its measurements, leap back into the air, and fly home.
Imagine a marine biologist standing on a beach. Instead of launching a boat, they throw this 250-gram robot into the air.
- The robot flies 3 miles out to a pod of whales.
- It dives into the water to record audio or take water samples.
- It launches back into the sky.
- It flies back to the researcher's hand to upload the data.
This could be done multiple times a day at virtually zero cost. It would allow us to monitor fragile marine ecosystems with a frequency and precision that is currently impossible.
The Reality Check on Autonomous Flight
We are not quite there yet.
While the physics of the FAAV are proven, the robot is not yet autonomous. During the tests in Lake Geneva, the robot was running on pre-programmed timing sequences. It did not have onboard sensors telling it where the water surface was, nor could it steer itself dynamically to avoid obstacles or handle sudden gusts of wind.
The researchers also have not yet linked the entire cycle together in a single, uninterrupted mission. They have proven the robot can dive, and they have proven it can launch itself out of the water. Putting those two halves together into a seamless, self-directed loop is the next big hurdle.
Additionally, the wings currently only flap up and down. They cannot twist or turn, which limits the robot's maneuverability. The team is already working on a new wing design that allows for active rotation, which will help the robot carve tight turns underwater and slice through turbulent, choppy winds in the air.
If you are looking to build or design bio-inspired hardware, the lesson here is clear. Stop trying to build complex mechanical joints to solve environmental transitions. Look at the passive properties of your materials first. Let the physics of the water do the bending for you, and keep your design as light and simple as possible.
