Using the same basic physics based on the familiar hollow tone produced by blowing across an empty bottle, researchers in Switzerland have developed a sound-powered propulsion system that can steer microrobots and even lift tiny machines.

In a new study published in Science Advances, engineers at the Swiss Federal Institute of Technology Lausanne (EPFL) demonstrated that 3D-printed cavities known as Helmholtz resonators can convert sound into directed jets of air and usable thrust.

The researchers built devices ranging from centimeters down to micrometers, used sound to propel and steer tiny boats, and created two microscopic flying machines—one microrobot lifted directly by acoustic thrust and another using rapidly spinning rotor blades. One version generated nearly five times its own weight in thrust, while another spun at up to 13,000 revolutions per minute to produce aerodynamic thrust.

The work points to a new way to power small robots without conventional motors, while also exposing the considerable engineering problems that remain before such machines can operate autonomously.

“While acoustic resonators have been extensively studied in liquids for the realization of untethered microrobots, their implementation in air remains largely unexplored,” researchers write. “The actuation scheme is compatible with even further miniaturization, pushing the boundaries of robotics and aeronautics.”

The idea behind the technology is surprisingly familiar. Blow across the top of an empty bottle and you hear a deep, hollow note created as the air trapped inside begins to vibrate.

Known as Helmholtz resonance, the effect happens because the air inside the bottle and its narrow opening behave somewhat like a mass attached to a spring. When sound at the right frequency excites a specially designed resonator, those vibrations become much stronger, forcing air rapidly through the opening. The resulting airflow can form a directional jet powerful enough to generate thrust.

The researchers set out to harness the same physics behind Helmholtz resonance as an unconventional propulsion method for microrobots.

The researchers designed tiny cavities that take advantage of the same principle. When sound hits one of these resonators at just the right frequency, the air in its narrow neck begins moving rapidly back and forth.

Because that airflow is slightly uneven as it enters and exits, it creates a small but focused jet of air. That jet produces thrust, giving the resonator enough force to help move or even lift a microrobot.

The team 3D-printed resonators ranging from centimeter to micrometer scales, designing them to respond to frequencies between roughly 200 hertz and 40 kilohertz. Experiments showed that larger cavities generally produced greater thrust, while the geometry and thickness of the resonator’s neck played a particularly important role in determining performance.

Importantly, researchers could tune different resonators to different frequencies. That means sound does more than supply power; frequency can effectively serve as a control signal.

The researchers demonstrated the concept using miniature boats equipped with acoustic resonators. One early prototype carried a resonator tuned to 1.2 kilohertz and moved only when exposed to sound at that frequency.

More advanced versions used separate resonators for propulsion and steering, allowing the researchers to guide a boat around obstacles remotely and along a figure-eight course with a mean tracking error of just 5.70 millimeters.

The team also showed that an external speaker is not always necessary. By attaching small vibration transducers directly to resonators, they could send sound through the robot’s structure itself.

This approach allowed the researchers to build a fully integrated boat equipped with its own battery, control electronics, and acoustic transducers. Bluetooth commands activated resonators tuned for propulsion and steering, with a total communication latency of about 30 milliseconds. The boat successfully traced the letters “EPFL,” navigated past obstacles, and could be steered back on course after researchers deliberately disturbed it by blowing air or physically pushing it.

However, shrinking the technology into flying microrobots provided the study’s most striking demonstration.

One microflier consisted of three microscopic resonators operating at 40 kilohertz and weighed only about 150 micrograms. Focused ultrasound caused each resonator to eject a tiny jet of air downward. Together, the three produced a thrust-to-weight ratio of approximately 4.9. The machine accelerated upward, climbed roughly 12 millimeters before settling around seven millimeters above its starting position.

A second design took a different approach. Instead of pointing the acoustic jets downward, the researchers used them to spin a rotor fitted with tiny aerodynamic blades. The entire aircraft weighed 184 micrograms.

Once exposed to ultrasound, the rotor accelerated to approximately 13,000 rpm. Its blades generated enough aerodynamic thrust for takeoff before settling to roughly 6,500 rpm while hovering.

Turning the device upside down prevented it from ascending, helping confirm that aerodynamic thrust, rather than acoustic radiation pressure alone, enabled flight.

Researchers demonstrated unconstrained flight with both microrobot designs. The direct-thrust version climbed rapidly but sometimes flipped in the air. The spinning version was more stable, apparently because its swift rotation provided a gyroscopic stabilizing effect similar to that found in helicopters and spinning discs.

Ultimately, the experiments remain proof-of-concept demonstrations rather than prototypes of free-ranging autonomous microrobots. Airborne propulsion requires external acoustic hardware capable of keeping a sufficiently strong sound field focused on a tiny moving object. That essentially limits the robot’s operating area. Putting the acoustic source onboard offers greater autonomy but might significantly increase the robot’s weight.

The current flying system also operates at a single 40-kilohertz frequency, restricting how independently its resonators can be controlled. The researchers suggest future multifrequency ultrasonic arrays could selectively activate different resonators, allowing small aircraft to steer and coordinate their movement through sound alone.

Further miniaturization is technically possible. Modern 3D nanoprinting can fabricate structures with resolutions around 700 nanometers, possibly allowing resonators only several micrometers across. However, at those scales, viscous forces become increasingly important and may dramatically reduce thrust, meaning simply shrinking the existing design may eventually stop working.

Still, the wider concept could go beyond flying machines. Researchers envision flexible structures containing many differently tuned resonators that can bend, vibrate, or change shape depending on the applied sound frequency, possibly resulting in shape-morphing microrobots and mechanical metamaterials.

“Our work extends the concept of acoustic manipulation toward active propulsion by harnessing cavity resonance to generate directional thrust and controlled motion,” the researchers write. “This approach broadens the scope of airborne acoustic actuation toward small-scale robots capable of autonomous operation in open environments.”

The recent study, “Acoustic Resonators as Wireless Actuators in Air for Small-Scale Robots,” was published in Science Advances. 

Tim McMillan is a retired law enforcement executive, investigative reporter and co-founder of The Debrief. His writing typically focuses on defense, national security, the Intelligence Community and topics related to psychology. You can follow Tim on Twitter: @LtTimMcMillan.  Tim can be reached by email: tim@thedebrief.org or through encrypted email: LtTimMcMillan@protonmail.com 

In a University of Kentucky laboratory, three small spacecraft stand-ins slid across low-friction air tracks, nudging and pulling one another without thrusters, propellant, or physical contact.

Instead, they used carefully timed magnetic fields, a technique that could one day help future satellite swarms hold precise formations in space without constantly burning fuel.

The NASA and Air Force-backed study, set to be published in the September 2026 edition of Aerospace Science and Technology,  marks a step toward what researchers call electromagnetic formation flying, a proposed method for controlling the relative positions of satellites using onboard coils rather than conventional propellant-based thrusters.

In the study, researchers demonstrated a decentralized, three-satellite control system that used alternating magnetic field forces to maintain formation with centimeter-level precision, offering a possible pathway for future space telescopes, interferometers, gravity-wave detectors, and other distributed spacecraft missions.

“Spacecraft formation flying can advance a variety of space technologies such as distributed-aperture telescopes, gravity-wave detectors, and interferometers,” researchers write. “One challenge for spacecraft formation flying is that traditional propellant thrusters eventually deplete, and they can contaminate sensitive spacecraft components.”

The study, authored by scientists from the University of Kentucky’s Department of Mechanical and Aerospace Engineering, focuses on a longstanding problem in spacecraft engineering.

Many future missions may not rely on a single large spacecraft, but instead on groups of smaller satellites flying in carefully arranged formations. If those satellites can maintain their spacing and orientation with extreme precision, they could function together like a much larger scientific instrument.

For example, a distributed-aperture telescope could use multiple spacecraft separated by carefully controlled distances to achieve observing capabilities that might be difficult or impossible with a single monolithic telescope.

Similar formation-flying concepts have been proposed for space-based interferometers and additional missions in which the relative positions of spacecraft are central to mission performance.

However, keeping multiple satellites in formation is not easy. Standard spacecraft rely on thrusters, which consume limited propellant. Over time, that fuel runs out, limiting mission duration and maneuverability. Thruster plumes can also contaminate sensitive instruments, a concern on spacecraft carrying delicate optical, scientific, or electromagnetic sensors.

However, electromagnetic formation flying, or EMFF, offers a different possibility. Each spacecraft carries electromagnetic coils. When electric current flows through those coils, they create magnetic fields. Those fields interact with magnetic fields created by nearby spacecraft, producing attractive or repulsive forces that can alter the satellites’ positions relative to one another.

Put simply, the satellites can push or pull on each other using magnetism.

The idea is not entirely new. Previous work has explored electromagnetic formation flying between two spacecraft, including experiments using direct-current-driven coils.

The problem becomes much harder when the formation includes three or more satellites. In that case, every satellite’s magnetic field can interact with the others, creating a web of coupled forces that becomes increasingly difficult to control as the formation grows.

The new study approaches that problem using alternating magnetic field forces, or AMFF. Rather than driving the coils with steady direct current, the researchers use sinusoidal currents at specific frequencies.

The key insight is that two satellites generate a nonzero time-averaged interaction force only when their magnetic moments alternate at the same frequency. If they do not share a frequency, the average force between them is negligible.

That frequency-matching approach effectively allows the system to pair satellites electromagnetically. One pair can interact at one frequency, while another pair interacts at a different frequency.

By adjusting the amplitude of each sinusoidal signal, the satellites can control the strength of the force between each pair without requiring a centralized controller to solve the entire formation’s motion at once.

“Only pairs with a common frequency can generate a nonzero time-averaged intersatellite force, and that force is determined solely by the amplitudes at the common frequency,” researchers write. “The method in this paper is decentralized, which means that each satellite has access to measurements of its position and velocity relative to only its local neighbor satellites.”

Centralized control can work in smaller systems, but it can turn computationally clumsy as satellite formations grow. A scalable system would need each spacecraft to make local decisions based on nearby neighbors, rather than requiring a single controller to calculate every force across the entire formation.

To test the idea, the researchers built a ground-based experimental platform using three custom electromagnetic actuation system units placed on linear air tracks.

The air tracks allowed the units to move with low friction, simulating one-dimensional relative motion between satellites. Each unit included an electromagnetic coil, batteries, power electronics, an Arduino Due microcontroller, wireless communication hardware, and laser range-finding sensors for measuring relative position.

In the three-satellite setup, the center unit was driven by a sum of two sinusoidal currents, while each outer unit was driven by one of those frequencies. This allowed the center unit to interact independently with each outer unit, demonstrating the central premise of frequency-multiplexed magnetic control.

In closed-loop experiments, the three-satellite system achieved formation in less than 30 seconds. The maximum steady-state formation error was less than plus or minus 0.01 meters, or about 0.4 inches. The mean steady-state formation error was less than plus or minus 0.001 meters, or about 0.04 inches.

Those numbers demonstrate that the method did more than produce a magnetic tug in a lab. It allowed multiple units to settle into a desired formation using local feedback and alternating magnetic fields, while avoiding the coupling problem that has limited electromagnetic control for larger formations.

The researchers also compared the experimental results with computer simulations. While the simulations generally matched the observed behavior, the team noted some differences caused by real-world hardware factors, including sensor interference, current saturation, possible misalignment, airflow disturbances along the air tracks, and unmodeled magnetic effects.

Those caveats are important. This was not a spaceflight demonstration, and the system was tested only in one dimension. Real spacecraft would operate in a far more complex environment, where microgravity, radiation, extreme temperature swings, atmospheric drag in low Earth orbit, gravitational forces, Earth’s magnetic field, and equipment restrictions would all matter.

Researchers also note that traditional thrusters may still be needed for orbit maintenance, orbit transfers, or controlling the overall center of mass of a satellite formation.

In essence, this is not a replacement for all spacecraft propulsion. It is better understood as a possible propellantless method for controlling the relative positions of satellites once they are already in orbit and operating as part of a formation.

Still, that limitation does not diminish the possible impact. If electromagnetic formation flying can be scaled and adapted for real spacecraft, it could reduce propellant demands for future distributed missions. That could extend mission lifetimes, reduce contamination hazards to sensitive instruments, and enable spacecraft formations more flexible than today’s architectures.

The research was supported in part by the National Aeronautics and Space Administration (NASA), the National Science Foundation, and the Air Force Office of Scientific Research.

The funding mix reflects the dual scientific and strategic interests in the work. NASA has clear reasons to explore precise formation flying for future space observatories and scientific missions. The Air Force, meanwhile, has an obvious interest in satellite coordination, maneuverability, and resilient distributed space systems.

The study’s most intriguing contribution may be its demonstration that magnetic-field-based satellite control can be both decentralized and scalable.

Rather than treating a multi-satellite formation in a single complicated system that must be controlled from the top down, AMFF allows each satellite to manage interactions with its neighbors using assigned frequencies. That could become increasingly valuable as space missions move toward constellations, swarms, and distributed architectures.

Researchers caution that additional work will be needed to move beyond one-dimensional air-track experiments and toward three-dimensional testing, more realistic sensing and control hardware, and, eventually, on-orbit validation.

Additional coils would be needed for three-dimensional position control, and relative attitude control could require additional actuation frequencies for each interacting satellite pair.

Nevertheless, the laboratory results show that the underlying concept works in a three-body system, which is exactly where the hard part begins.

Two satellites can pull or push on each other relatively easily. Three satellites introduce the coupling problem that makes electromagnetic formation flying difficult. By demonstrating three-satellite control with alternating magnetic fields, the study delivers experimental evidence that the frequency-based solution can work.

For forthcoming space missions, that could be a meaningful step. The next generation of telescopes and orbital instruments may not always be built as single enormous spacecraft. They may be assembled virtually, with multiple satellites flying together as one coordinated machine.

If so, keeping those satellites precisely arranged without wasting propellant could become one of the enabling technologies of future astronomy, Earth observation, and national security space systems.

“This article presented a 3-satellite experimental demonstration of decentralized EMFF using AMFF,” researchers conclude. “The settling time and steady-state formation error achieved in these experiments demonstrate that this approach is most likely capable of meeting requirements for satellite constellation reconfiguration and maintenance, thereby addressing the need for propellant-free formation flying.”

Tim McMillan is a retired law enforcement executive, investigative reporter and co-founder of The Debrief. His writing typically focuses on defense, national security, the Intelligence Community and topics related to psychology. You can follow Tim on Twitter: @LtTimMcMillan.  Tim can be reached by email: tim@thedebrief.org or through encrypted email: LtTimMcMillan@protonmail.com