MIT and EPFL Team Up to Build a Robo-Puffin: A Robot That Moves Through Air, Water, and Sea

When a Robot Learns to Cross Boundaries Like a Puffin

Water is about 1,000 times denser than air, and moving efficiently through both media typically requires radically different mechanical structures. But nature already has a solution: birds such as gulls and puffins can dive into water at high speed, swim underwater, and then leap back into the air and fly away. Engineers at MIT and the Swiss Federal Institute of Technology in Lausanne (EPFL) drew inspiration from these amphibious birds to develop a robot that can fly through the air, glide on the water surface, and swim underwater. It closely resembles a puffin, earning it the nickname "robo-puffin."

This design breaks away from the conventional "single-medium" mindset of traditional robots, pointing toward a more versatile cross-medium locomotion platform.

Why Is Crossing Media So Difficult?

The physical properties of air and water differ enormously: water is roughly 1,000 times denser than air and about 50 times more viscous. Aircraft pursue lightweight wings and low-drag shapes, while underwater vehicles require pressure-resistant hulls and efficient propulsors. Most cross-medium approaches adopt a "morphing" strategy—such as foldable wings or retractable propellers—but these come at the cost of structural complexity and weight.

The puffin's solution is "one tool, many uses": short, stiff wings provide lift in the air and switch to paddling strokes underwater; a streamlined body reduces water-entry impact; and the composite structure of feathers and bones balances buoyancy control with pressure resistance. The MIT and EPFL team sought to reproduce this multifunctional integrated design in engineering terms, rather than simply stacking two separate systems.

Another key physical obstacle in cross-medium locomotion is water-entry impact loading. When an object crosses a free surface at high speed, it experiences severe transient pressure peaks. This impact load is proportional to the square of the entry velocity and is closely related to the entry angle. For an aerial vehicle, this impact is enough to destroy lightweight structures, so birds like puffins retract their wings before entry and cut into the water at a specific angle, distributing the impact across their streamlined bodies. In engineering terms, this requires the robot to have variable sweep angles or folding mechanisms, while ensuring that the hull does not undergo buckling instability at the moment of entry. In addition, water adhesion is particularly significant during the exit phase: as an object leaves the water, surface tension and cavity collapse create additional drag—equivalent to adding another effective weight to the robot itself. This is the fundamental reason why leaping out of water requires explosive thrust.

How the Robot Achieves Three-Mode Locomotion

Aerial Flight

According to the original report, the robot retains fixed-wing flight capability. A puffin's wings flap at high frequency during flight, while the robot may use flapping-wing or fixed-wing gliding approaches. Specific aerodynamic parameters were not disclosed in the summary, but the team emphasizes that its aerial maneuverability is sufficient to support spotting a target from the air and diving into the water.

Water Entry and Underwater Swimming

The impact force at the moment of water entry is a key challenge for cross-medium robots. Puffins reduce impact by retracting their wings and adjusting their entry angle. The robot must balance structural strength against weight while ensuring that the propulsion system is not damaged by water pressure after entry. The original material mentions that it can "plunge into the sea, swim underwater," indicating that water entry and underwater propulsion have been physically validated.

Water Exit and Re-Takeoff

Leaping from water back into the air is the hardest step: the robot must overcome water adhesion and its own weight to gain sufficient exit velocity and attitude. Puffins achieve the leap through powerful wing slaps on the water and webbed-foot pushes. In engineering terms, this may require a combination of explosive propulsion and rapid attitude adjustment. The original report does not provide details on successful water exit, but "leap back into the air to fly away" implies that this phase has been demonstrated.

Achieving stable flight after leaping out of the water also involves the problem of cross-medium control switching. Underwater, the robot relies on buoyancy and hydrodynamic forces for attitude control; at the moment of exit, it must switch to aerodynamic control laws. The sensor validity and actuator response bandwidth of the two are completely different. More challenging still, during exit the robot may undergo a continuous transition from fully submerged to partially submerged to fully emerged, during which hydrodynamic coefficients change dramatically. The controller must maintain stability under such large-range parameter perturbations. Biological puffins rely on rapid fusion of vision and proprioception to cope with this transition, while in engineering it is necessary to design robust mode-switching logic or a unified reference model to avoid attitude instability at the moment of exit.

Design Philosophy: Learning Integrated Design from Amphibious Birds

Traditional multimodal robots often make mechanical switches between "flight mode" and "swimming mode," introducing extra weight and failure points. The MIT/EPFL approach is closer to biological integrated design: the same propulsion structure changes how it operates in different media, rather than being wholly replaced.

The potential advantages of this approach are a more compact structure, lighter weight, and faster mode switching. The trade-off is that efficiency in each individual mode may be slightly lower than that of a dedicated platform. For scenarios requiring frequent transitions between water and air (such as ocean monitoring and search-and-rescue), this compromise may be worthwhile.

Potential Applications and Open Questions

The most direct application scenarios for this type of cross-medium robot include coastline patrol, marine biological observation, water search-and-rescue, and missions requiring rapid arrival at the water surface from the air followed by underwater diving. Compared with deploying drones and underwater vehicles separately, a single three-mode platform can reduce deployment steps and shorten response time.

However, the original information is only the beginning of the report, and many key metrics have not yet been disclosed: endurance, underwater dive depth, payload, level of autonomous control, and the success rate of water exit and re-takeoff. These data will determine how far the robot is from practical use.

Editorial View

Cross-medium robots are not a new concept, but most approaches remain at the laboratory proof-of-concept stage and often sacrifice performance in one medium. If the MIT and EPFL "robo-puffin" can achieve three-mode locomotion while maintaining structural simplicity, its value lies not in leading any single metric but in demonstrating the engineering feasibility of biological integrated design. What is worth watching next is whether the team publishes complete locomotion data and autonomous control solutions—that is the watershed between "can fly and can swim" and "useful."

One-line Summary

MIT and EPFL, inspired by puffins, developed a cross-medium robot that can fly, swim, and take off again.

Paragraph Summary

Engineers at MIT and EPFL drew inspiration from amphibious birds such as puffins to develop a cross-medium robot that can fly through the air, glide on the water surface, and swim underwater, closely resembling a puffin. Water and air differ in density by about 1,000 times, and traditional cross-medium approaches rely on morphing structures such as foldable wings or retractable propellers, at the cost of increased weight and complexity. This robot instead borrows from biology's integrated design: the same propulsion structure changes how it operates in different media. Water-entry impact, water-exit leaping, and mode-switching efficiency are the three major engineering challenges. Potential applications include coastline patrol, ocean monitoring, and water search-and-rescue, but key metrics such as endurance, dive depth, and level of autonomous control have not yet been disclosed.