Robot swims and flies like diving birds, could revolutionize ocean sampling

Engineers at EPFL's School of Engineering and MIT have created a novel robot inspired by diving birds—species like loons, gulls, puffins, and petrels that plunge into water to hunt and then leap back into the air to fly. The "flapping-wing aerial-aquatic vehicle" (FAAV) weighs less than 300 grams and was designed to help scientists understand how diving birds adapt their flight mechanics to move through air and water, which have very different physical properties.

The robot's design includes a central fuselage, two flexible flapping wings made of thin membranes coated with hydrophobic nanoparticles to repel water, and a motorized steerable tail. A waterproof electric motor drives a crankshaft that pumps the wings up and down at preset frequencies. The wings and tail can be swapped for different sizes to explore various configurations. Raphael Zufferey, lead author and former EPFL postdoctoral fellow now an assistant professor of mechanical engineering at MIT, began the work under the supervision of EPFL lab heads Dario Floreano and Auke Ijspeert. He completed the research at MIT, where he now leads the AURA Lab.

Testing in a water tank at EPFL and in Lake Geneva revealed critical parameters for successful air-water transition. The robot could swim at speeds of almost one meter per second when flapping at around 5 hertz (five flaps per second), and fly at around 6 meters per second at similar flapping frequencies—speeds comparable to actual diving birds. Wing size (80 centimeters) and flexibility proved essential; wings must flex enough to minimize flapping amplitude underwater yet remain rigid enough to keep the robot aloft in air. Crucially, to launch from water into air, the robot must be pitched at a steep 70-degree angle to prevent its wingtips from touching the water's surface.

A surprising finding was that the robot could achieve flight without feet—a capability most diving birds lack. "If you look at birds, most birds need to paddle their feet at the surface to take off," Zufferey explained. "And the question was, do we need the same for robots? And it turns out we don't. No one's been able to fly out of the water with wings." This wing-only solution simplifies the design and opens possibilities for new aerial-aquatic vehicles. Looking ahead, the team plans to improve wing design to enable turning in addition to flapping, test performance in turbulent conditions such as choppy waters and wind, and eventually deploy the robot to help oceanographers, marine biologists, and coastal communities study hazardous aquatic regions—icebergs, whale pods, port facilities—at a fraction of the cost of traditional ocean vessels. Results were published in Science.

The robot represents a convergence of biomimetic engineering and practical ocean science. Engineers have long studied diving birds—about 100 species can both swim and fly—to understand how animals move through two physically distinct fluids. Air and water have vastly different densities and viscosity, so wings that work in one medium must be carefully adapted for the other. By testing combinations of wing size, flapping frequency, and tail angle in a water tank and Lake Geneva, the team discovered that an 80-centimeter wing with around 5 hertz flapping frequency and a 70-degree pitch angle allows smooth transitions from water to air.

A key insight emerged from comparing the robot's design to its biological inspirations. Most diving birds use their feet to paddle at the surface before taking flight; the EPFL–MIT team found that their wing-only design eliminates this requirement. This difference is significant for engineering because it simplifies the robot's mechanics and reduces moving parts, making it more durable and easier to control. The hydrophobic wing coating helps shed water, mimicking adaptations in real diving birds.

The practical motivation is clear: oceanographers and marine biologists struggle to sample hazardous aquatic environments—floating ice, whale pods, isolated facilities—without risking expensive research vessels. A small, autonomous, wing-flapping robot that can launch from a boat, fly to a target, dive for samples, and return could dramatically reduce cost and risk. The researchers are now refining the design to allow the wings to turn as well as flap, and to validate performance in realistic choppy waters and wind before deploying the robot for real ocean science missions.

No comments yet. Be the first to share your thoughts!