Engineers at a Massachusetts Institute of Technology laboratory have developed a robotic vehicle costing just $300 that can fly through the air and swim underwater. The device was designed to navigate complex environments by mimicking the physical movements of diving seabirds.
The vehicle, named the Flapping-wing Aerial-Aquatic Vehicle, or FAAV, uses flexible flapping wings to propel itself through both air and water. Lead researcher Raphael Zufferey, an assistant professor of mechanical engineering at MIT, said the team hopes the robot will eventually be deployed to monitor and help protect marine ecosystems.
The Massachusetts Institute of Technology, based in Cambridge, Massachusetts, is widely recognized for its pioneering research in autonomous systems and mechanical engineering. Designing vehicles capable of operating in both air and water has long presented a major challenge in robotics due to the contrasting physical properties of the two environments.
Inspiration from diving seabirds
To design the robot, the research team studied petrels, puffins, and other diving birds that use their wings for propulsion both in the sky and below the surface of the water. The specialists in aero-aquatic robotics analyzed the complex fluid dynamics that allow these birds to move seamlessly between the two environments.
Zufferey noted that although previous studies had examined the mechanics of diving bird wings, no researchers had previously figured out how to translate those physical principles into a fully mobile robot. The team reviewed all available scientific literature on diving birds, mapping how wingbeat frequencies varied in air and water across species with different wingspans.
Their analysis confirmed that larger birds exhibit lower wingbeat frequencies, providing a blueprint for sizing and powering the robotic platform.
Engineering trade-offs and wing design
The resulting prototype weighs 250 grams, powered by a small battery-driven motor that operates flapping nylon wings and a tail coated in water-repellent nanoparticles. Despite its bird-like appearance, Zufferey emphasized that the FAAV is not a strict copy of biological systems.
Water is hundreds of times denser than air, with factors such as water temperature, pressure, and air humidity altering the physical forces acting on the vehicle. Living diving birds partially fold their wings underwater to reduce wing stroke amplitude, which is the vertical distance a wing travels during a single stroke, thereby minimizing drag while generating forward movement.
Replicating a folding wing mechanism in a robot would have required four additional mechanical joints, adding significant engineering complexity and vehicle mass. Instead, the MIT team developed a rigid wing that does not fold but possesses greater flexibility than a bird wing to restrict stroke amplitude underwater.
The control system on the FAAV does not detect whether the robot is moving through air or water. Instead, the vehicle is programmed to maintain a target wingbeat frequency per second regardless of the surrounding medium.
Speed, range, and field testing
In aerial flight, the FAAV reaches speeds slightly exceeding 6 meters per second, equivalent to 13.4 miles per hour. Underwater, the robot swims at nearly 1 meter per second, or 2.2 miles per hour.
Theoretical calculations indicate the robot could cover 6 kilometers, or 3.7 miles, in the air, or swim 2 kilometers, or 1.2 miles, on a single battery charge. Researchers noted that these maximum range figures have not yet been tested in field conditions.
The development team spent a year testing the FAAV in a specialized water tank in Massachusetts before conducting open-water trials at Lake Geneva in Switzerland. Much of the fine-tuning involved establishing the optimal angle for the robot to dive into the water, which researchers determined to be 70 degrees, as well as perfecting how it climbs out of the water back into flight.
Zufferey stated that the vehicle can navigate moderate wind and wave conditions, although the current prototype cannot operate in severe weather. Lake Geneva, a large freshwater lake shared by Switzerland and France, provided researchers with controlled open-water conditions to evaluate how the robot handled real-world wave action.
Independent evaluation and operational challenges
Maaten Furlong, the director of engineering science at the National Oceanography Centre in the United Kingdom, who was not involved in the study, praised the technical achievement. Furlong stated that creating a single vehicle capable of operating effectively in both air and water represents a major engineering hurdle, calling the integration of both modes an impressive accomplishment.
However, Furlong raised questions regarding the robot's payload capacity, structural durability, regulatory compliance, and overall viability in open-ocean environments. He noted that combined aerial and underwater vehicles remain relatively unexplored because they are technically difficult to design and their operational advantages have not yet been clearly demonstrated.
Furlong added that few bio-inspired robots progress from academic laboratories to practical operational deployment. In practice, traditional electric motors and propellers are highly efficient, cheap, and reliable, making them difficult to surpass in terms of cost and performance.
Despite those reservations, Furlong said oceanographers are generally technology agnostic, focusing less on how measurement tools operate and more on obtaining reliable, high-quality data at low cost. He noted that if the flapping-wing platform can demonstrate those operational attributes, marine scientists are likely to adopt it.
Oceanographic sampling and future development
The MIT research team is currently seeking funding and operational partnerships to further advance the vehicle. Zufferey explained that while the published study proved that flying, swimming, diving, and transitioning between mediums are all possible individually, the team has not yet integrated all those capabilities into a single autonomous mission.
The researchers hope the FAAV will serve as a low-cost tool for oceanographic data collection. Gathering scientific samples at sea using traditional research vessels and specialized submersibles remains extremely expensive.
Zufferey suggested that the lightweight robot could be launched from land or ships, fly autonomously along pre-programmed flight paths, dive underwater to collect samples, and return to base. He highlighted potential deployment scenarios in hazardous environments, such as collecting water samples from toxic algal blooms, volcanic lakes, or areas adjacent to melting icebergs.
The vehicle could also be outfitted with miniaturized cameras to monitor marine wildlife without causing disruption. Zufferey calculated that by utilizing carbon fiber and other lightweight composite materials, the underlying flapping-wing design could be scaled up to a wingspan of up to 15 meters, or 49 feet.
At its current scale, the component parts for the FAAV cost $300. Zufferey estimated that even with a more powerful motor and a reinforced chassis, total construction costs would not exceed $1,000, which he described as extremely cheap for ocean science applications. Zufferey has since set up a new laboratory at MIT featuring a larger water tank and high ceilings to support expanded testing.
