Could Acoustic Levitation Be Possible Today

Ask most people what it would take to make an object float in place, unsupported, in the middle of a room, and they will describe something out of science fiction: magnetic rails, antigravity plating, or a barely explained “force field.” Ask a physicist who has spent time in an ultrasonics lab, and the answer is far more mundane and far more immediate. You do not need to defeat gravity to make something hover. You only need to push against it hard enough, in the right pattern, with something that is already all around us: sound.

Acoustic levitation is not a laboratory curiosity from a distant future. It is a mature, well-understood, and increasingly accessible technology that already exists on bench tops in university labs, in pharmaceutical research facilities, in materials-science clean rooms, and even in hobbyist workshops built around parts that cost less than a video game console. The physics has been worked out since the 1930s. The hardware needed to generate the necessary sound fields has become cheap, small, and programmable only in the last decade and a half, thanks to advances in piezoelectric transducers, digital signal processors, and phased-array electronics originally developed for entirely different industries such as telecommunications and medical ultrasound. What used to require a specialized acoustics laboratory with custom-machined horns and high-voltage amplifiers can now be built on a 3D-printed frame with parts ordered from an electronics supplier for the price of a modest dinner out.

This raises the central question this article is built around: how, exactly, could acoustic levitation be made to work today, using tools, materials, and physical principles that are already available, rather than ones that remain speculative or decades away? The honest answer is that in many senses it already does work today — acoustic levitation is not a hypothetical, it is a demonstrated and repeatable phenomenon performed daily in laboratories around the world. But the deeper and more useful question, the one this article will answer in detail, is how the underlying physics translates into practical, buildable systems: what generates the lifting force, what determines the size and mass of objects that can be suspended, what hardware architectures exist, what has changed recently to push the boundaries of what is possible, and what a person — whether a curious hobbyist, an engineering student, or an industrial researcher — would actually need to assemble a working levitator with today's technology.

As someone whose work sits at the intersection of gravitational physics and the practical engineering of forces that can counteract or locally mimic the effects of gravity, I find acoustic levitation fascinating precisely because it is not a trick of illusion or a niche party demonstration. It is a genuine, quantifiable balancing of forces: the downward pull of gravitational acceleration against the upward push of acoustic radiation pressure, mediated by nothing more exotic than compressions and rarefactions in ordinary air. Understanding how that balance is achieved — and how it has become dramatically easier to achieve in the last several years — is the subject of everything that follows.

This article proceeds in stages. First, it lays out the physical principles that make acoustic levitation possible at all: the nature of sound as a pressure wave, the concept of acoustic radiation force, and the mathematics that governs when and how an object can be trapped in a sound field. Second, it traces the historical development of the technology, from its theoretical origins in the 1930s through the breakthroughs of the 1970s, 1980s, and the more recent explosion of phased-array and holographic techniques. Third, and most practically, it examines the hardware that exists today — transducers, drivers, control electronics, and array architectures — that make building a levitator feasible with off-the-shelf components. Fourth, it surveys what has actually been levitated, from water droplets and insects to two-centimetre polystyrene spheres and biological cell cultures, and explains the physical limits on size, density, and mass that any levitation attempt must respect. Fifth, it walks through the actual applications being pursued today in pharmaceuticals, materials science, biomedicine, electronics manufacturing, and space research. Sixth, it addresses the real limitations, safety considerations, and open engineering challenges that anyone attempting to build or scale up an acoustic levitation system will encounter. And finally, it looks at where the technology is heading, including very recent breakthroughs in multi-object manipulation that point toward genuinely new capabilities.

Throughout, the emphasis is on what is achievable now, with parts, knowledge, and manufacturing capability that already exist — not on speculative future technology. Acoustic levitation is one of the rare domains where the gap between “theoretically possible” and “practically achievable on a budget” has nearly closed. That closing gap is the real story.

The Physics of Sound as a Levitating Force

To understand how sound can hold an object against gravity, it helps to first discard the intuitive picture of sound as something that merely tickles the eardrum. Physically, sound is a mechanical wave: a travelling pattern of compression and rarefaction in a medium, whether that medium is air, water, or a solid. As a sound wave propagates through air, it does not simply carry information about pitch and loudness — it carries momentum. Every region of compressed air pushes on the air in front of it slightly harder than the rarefied air pushes back, and while this imbalance is minuscule for the sound levels we experience in everyday conversation or music, it becomes significant, and eventually dominant, at high intensities and high frequencies.

This momentum transfer is called acoustic radiation pressure, and it is the direct acoustic analog of the radiation pressure exerted by light, the same phenomenon that pushes on a solar sail or, at a much larger scale, shapes the tails of comets. The crucial difference is that sound waves, unlike light, require a medium to propagate, and the forces involved, watt for watt, are vastly larger than those achievable with light in air. This is why optical tweezers — which use focused laser light to trap and move microscopic particles — are limited to manipulating objects on the scale of single cells or smaller, while acoustic tweezers can comfortably lift objects the size of grapes, dice, or small insects. The energy density achievable in an acoustic field, especially in air, produces forces per unit of input power that dwarf what is achievable optically, which is one of the central practical reasons acoustic levitation has become such an attractive technology for manipulating larger, everyday-scale objects without contact.

The forces at play in acoustic levitation arise from a nonlinear effect. If you look at a single, low-intensity sound wave, the pressure oscillates symmetrically above and below atmospheric pressure, and the time-averaged force on any object placed in that wave is essentially zero — the pushes and pulls cancel out. But at higher intensities, particularly when a standing wave is formed by reflecting sound back on itself, this symmetry breaks down. The interaction between an object and a high-intensity sound field produces a net, time-averaged force that does not cancel to zero. This residual force is the acoustic radiation force, and it can be strong enough, when properly directed, to counteract gravity entirely for a lightweight object.

A standing wave is central to nearly every practical acoustic levitator built today. When a sound wave travelling in one direction meets a wave of the same frequency travelling in the opposite direction — typically because the original wave has reflected off a surface and is now interfering with itself — the two waves combine to produce a stationary interference pattern. Instead of a wave that appears to move through space, you get a pattern that oscillates in place: certain points, called nodes, where the air pressure barely changes at all, and other points, called antinodes, where the pressure swings through its largest excursions. Between the nodes and antinodes, the acoustic radiation pressure gradient creates a restoring force that pushes small, light objects toward the pressure nodes and holds them there, resisting displacement in any direction, much like a ball resting in the bottom of a bowl resists being pushed uphill.

This is the essential trick of standing-wave acoustic levitation: by generating a periodic pattern of pressure nodes in the air using a source and a reflector (or, in modern designs, an array of sources acting as their own effective reflector), an engineer creates a series of stable equilibrium points suspended in midair, spaced at intervals of half the acoustic wavelength. An object placed near one of these nodes experiences a restoring force in all three dimensions if the field is properly shaped, and it will simply sit there, trapped by sound, as long as the driving signal remains on and the field geometry remains stable.

It is worth being precise about what “trapped” means physically. The object is not floating because sound is somehow lighter than air or because it is being “pushed up” in a simple, single-direction sense in the way a fan pushes a balloon upward. Rather, the object sits in a genuine three-dimensional potential well, similar in concept (though very different in physical origin) to the potential wells that trap ions in an ion trap or atoms in an optical lattice. Displace the object slightly from the node, and it experiences a force pulling it back. This is a stable equilibrium, not a precarious balancing act, which is why levitated objects in a well-designed acoustic trap can sit motionless for hours, and in demonstrated cases, for many hours at a stretch without any active feedback control.

Two more physical ingredients matter enormously in practice: frequency and intensity. Frequency determines the acoustic wavelength, and the wavelength in turn sets the natural size scale for everything about the levitator — the spacing between transducer elements, the spacing between pressure nodes, and critically, the maximum size of object that can be stably trapped using the simplest standing-wave technique. Most practical airborne acoustic levitators today operate at ultrasonic frequencies around 40 kHz, a frequency chosen not because it is uniquely magical but because it sits just above the upper limit of typical adult human hearing (roughly 20 kHz), it is a frequency at which small, cheap, mass-produced piezoelectric transducers are widely available (the same components used in ultrasonic distance sensors, parking sensors, and proximity detectors), and it produces a wavelength of about 8.6 millimetres in air at room temperature, which is a convenient size for lifting millimetre-to-centimetre-scale objects.

Intensity, meanwhile, determines how strong the trapping force is, which in turn determines how large, heavy, or dense an object can be levitated. The acoustic radiation force scales with the square of the pressure amplitude, meaning that doubling the sound pressure roughly quadruples the available lifting force. This is why serious acoustic levitators use multiple transducers driven in a coordinated, phase-matched way rather than a single small speaker: combining the outputs of dozens or hundreds of individual ultrasonic emitters, all firing in a phase-controlled pattern that constructively reinforces the field at the trapping location, allows the accumulated acoustic pressure — and therefore the lifting force — to be pushed far higher than any single transducer could achieve alone. This is precisely the insight that makes modern, multi-transducer levitator designs so much more capable than the single-emitter designs of decades past, and it is one of the central reasons acoustic levitation has become dramatically more practical in the last fifteen years.

Historical Development of Acoustic Levitation

The idea that sound could suspend matter against gravity is not a recent one, though the tools needed to do it well are. The theoretical foundation was laid in 1934 by the German physicist Rudolf Bücks and later, more comprehensively, by the Russian physicist Lev Gor'kov, whose 1962 formulation of the acoustic radiation potential remains the mathematical backbone of nearly every levitation calculation performed today. Gor'kov's contribution was to show that for a small particle in an arbitrary sound field, a time-averaged potential energy function could be derived that depended entirely on the particle's size, density, and compressibility relative to the surrounding medium, and on the local acoustic pressure and velocity fields. Wherever that potential has a local minimum, a particle placed nearby will be trapped. This turned acoustic levitation from a qualitative curiosity into a quantitative engineering problem: if you know the sound field, you can calculate exactly where an object will be held, how strongly, and under what conditions it will escape.

Practical demonstrations followed the theory by decades, constrained less by physics than by available technology. Early acoustic levitation experiments in the mid-twentieth century relied on Langevin-type transducers — the same basic piezoelectric horn technology developed originally for sonar during and after the First World War — driven at high voltages, often exceeding one hundred volts, and requiring careful mechanical tuning to hit a precise resonant frequency. These devices worked, but they were finicky. A Langevin horn had to be manufactured to tight tolerances, and its resonant frequency would drift as the transducer heated up during use, requiring constant retuning or active frequency-tracking electronics to keep the standing wave properly established. This made early acoustic levitators laboratory instruments in the truest sense: expensive, delicate, and confined to the hands of specialists.

The 1960s and 1970s saw the technology mature within a fairly narrow research niche, largely oriented around materials' science. Researchers realized that acoustic levitation offered something almost no other technique could: a way to hold a sample of molten metal, glass, or other high-temperature material in place without ever letting it touch a container wall. This “containerless processing” capability mattered enormously for materials science because container walls are a constant source of contamination and nucleation sites that disturb the natural solidification behaviour of a melt. If you want to study how a metal alloy solidifies in its purest, most homogeneous form, or how a liquid can be super cooled well below its normal freezing point without crystallizing, removing the container is transformative. Acoustic levitation, along with electromagnetic and aerodynamic levitation, became one of the standard tools for these kinds of containerless materials research, and remains so today in various specialized furnaces around the world.

A major turning point came with the involvement of space agencies in the 1980s and 1990s. Taylor Wang, a physicist working with NASA's Jet Propulsion Laboratory, led experiments using acoustic levitation as a containment method for fluid dynamics studies aboard the Space Shuttle, flying an acoustic levitation apparatus on the Challenger mission STS-51-B in 1985 to study the behaviour of levitated liquid droplets in microgravity. Follow-on experiments were conducted aboard the United States Microgravity Laboratory missions in 1992 and 1995. These flights were significant not just as scientific milestones but as proof that acoustic levitation techniques could be engineered reliably enough to trust with a costly, one-shot spaceflight experiment, and they cemented acoustic containment as a serious tool for studying fluid physics, combustion, and materials behaviour in environments where gravity itself was no longer confounding the results.

Through the 1990s and 2000s, acoustic levitation research diversified into biology and chemistry, where the containerless, contact-free environment proved useful for handling small volumes of reactive or precious samples, studying the drying and crystallization behaviour of levitated droplets, and observing biological specimens such as small organisms without the mechanical stress of physical handling. But the equipment remained expensive and specialized: high-voltage Langevin transducers, careful mechanical alignment, and dedicated laboratory space were still the norm.

The real inflection point for accessibility came in the 2010s, driven by two parallel developments. The first was the maturation of low-cost, mass-produced piezoelectric ultrasonic transducers, originally developed for entirely unrelated commercial applications such as automotive parking sensors, ultrasonic cleaning devices, and proximity sensing in consumer electronics. These transducers, often available for a fraction of a dollar each when purchased in bulk, could be driven at low voltages of around twenty volts rather than the hundred-plus volts required by traditional Langevin horns, and did not require the same painstaking mechanical tuning. The second development was the rise of cheap, capable microcontroller and digital signal processing hardware — inexpensive boards such as the Arduino and later dedicated field-programmable gate array and digital signal processor chips — that made it feasible to generate and precisely control the phase-matched driving signals needed to coordinate dozens or hundreds of individual transducers.

The convergence of these two trends is what produced the modern acoustic levitator, exemplified most clearly by the TinyLev design published by researchers Asier Marzo, Adrian Barnes, and Bruce Drinkwater at the University of Bristol in 2017. Rather than relying on a small number of large, high-voltage Langevin transducers, TinyLev used arrays of dozens of small, cheap, well-matched 40 kHz transducers of the type used in ultrasonic rangefinders, wired in parallel and driven by a simple, low-voltage circuit. The result was a levitator that could be built from off-the-shelf parts for a modest cost, assembled with the help of a 3D printer, and operated without any of the finicky tuning that had made earlier designs the exclusive province of specialized labs. This design point — many cheap emitters working in concert rather than a few expensive, precisely engineered ones — has become the dominant architecture for nearly all new acoustic levitation systems built since, from academic research platforms to hobbyist projects, and it is the single biggest reason acoustic levitation has moved from a laboratory rarity to something that can be reproduced on a kitchen table.

Alongside this hardware democratization, the last decade has also seen a parallel revolution in control sophistication. Where early standing-wave levitators could only hold an object at a fixed node determined by the physical geometry of the source and reflector, researchers at Bristol and elsewhere developed phased-array techniques that allow the acoustic field itself to be reshaped electronically, moving the trapping point in three dimensions, splitting a single trap into multiple independent traps, or generating exotic field shapes such as acoustic vortices and acoustic “bottle” traps that can hold objects in configurations impossible with a simple standing wave. This shift from fixed, mechanically defined traps to electronically programmable ones, borrowing conceptually from the earlier development of holographic optical tweezers in laser physics, is often referred to as holographic acoustic tweezing, and it represents the current frontier of what acoustic levitation hardware can do.

The Gor'kov Potential and the Mathematics of Acoustic Trapping

Understanding why acoustic levitation works at a level useful for actually building something requires a slightly closer look at the mathematics, though it can be kept largely conceptual without sacrificing the essential insight. The starting point is Gor'kov's radiation potential, which applies to particles that are small compared to the acoustic wavelength — a condition satisfied by most objects levitated in standard 40 kHz systems, since the wavelength at that frequency is around 8.6 millimetres and typical levitated objects (polystyrene beads, water droplets, small insects) are a few millimetres across or smaller.

Gor'kov showed that the time-averaged force on such a small particle can be derived from a scalar potential function that depends on two properties of the field — the mean-square acoustic pressure and the mean-square acoustic particle velocity, both averaged over a cycle — combined with two properties of the particle and surrounding medium: the compressibility contrast between the particle and the air (how much more or less easily the particle can be squeezed compared to the surrounding air) and the density contrast between the particle and the air. The potential has two competing terms: one driven by pressure that tends to push particles toward pressure nodes (a monopole scattering effect related to compressibility) and one driven by velocity gradients that tends to push particles toward velocity nodes, which for typical materials with positive density contrast against air, works in the same direction (a dipole scattering effect related to density). For nearly all solid and liquid objects levitated in air — which are almost always denser and less compressible than the surrounding air — both terms push the particle toward the same location: the pressure node of the standing wave, which is simultaneously a velocity antinode.

This is a critical and somewhat counterintuitive point worth dwelling on: objects levitated acoustically in air do not sit at the point of maximum pressure, as a naive mental picture of “sound pushing on an object” might suggest. They sit at the pressure node — the point where the time-averaged pressure barely fluctuates at all — but where the velocity of the air oscillation is at its maximum. The lifting mechanism is not a simple, constant “gust” of high-pressure air holding the object up, the way a leaf blower holds up a ping-pong ball. It is a genuine trapping potential with a well-defined minimum, arising from the second-order (nonlinear) interaction between the object and the oscillating field, and the restoring force grows the further the object is displaced from that minimum, exactly like a ball in a bowl.

For the specific and very common case of a standing wave formed between a source and a reflector, the geometry simplifies considerably. Pressure nodes occur at regular intervals spaced by half the acoustic wavelength along the axis between source and reflector, so a standing wave levitator naturally produces a stack of multiple potential trapping planes rather than just one, which is why it is common to see several small levitated objects stacked vertically, each hovering at a different node, in demonstrations of this type of levitator.

Two practical design consequences fall directly out of this mathematics, and both matter enormously for anyone actually trying to build a levitator. First, the maximum size of a stably trapped object is fundamentally constrained by the wavelength because the simple Gor'kov formulation assumes the particle is small relative to the wavelength; once an object approaches or exceeds the wavelength in size, the physics becomes considerably more complicated, scattering effects become nonuniform across the object's surface, and stability becomes much harder to achieve without more sophisticated field shaping — a limitation that, as described later in this article, has only recently been substantially overcome through clever multi-transducer geometries. Second, the trapping force scales with the acoustic pressure amplitude squared and with the volume of the particle, but the gravitational force the trap must overcome scales with the particle's mass, meaning that for objects of a given material, there is a maximum size beyond which gravity wins regardless of how strong the field is made, unless the field intensity is increased proportionally. This is why acoustic levitators are generally most comfortable holding low-density materials such as expanded polystyrene foam, and increasingly challenged as target density rises toward that of water, and more challenged still for genuinely dense materials such as metals, though even metal levitation has been demonstrated for small enough samples with sufficiently intense fields.

Near-Field, Far-Field, and Standing-Wave Architectures

Not all acoustic levitation systems work the same way, and understanding the differences between the major architectural approaches clarifies why certain designs are practical to build today while others remain confined to specialized research settings.

The most common and most easily replicated architecture is the standing-wave levitator, sometimes described as a near-field or resonance-based design when the reflector distance is tuned precisely to a resonant condition. In its simplest form, this consists of a single ultrasonic transducer (or a small cluster of them acting as one emitter) facing a rigid, sound-reflecting surface, typically a concave or flat metal or plastic reflector, positioned at a distance that is an integer multiple of half the acoustic wavelength. Sound emitted by the transducer travels to the reflector, bounces back, and interferes with the outgoing wave to create a stationary standing-wave pattern in the gap between them. This is the architecture used in the classic tabletop demonstration levitators seen in many introductory physics videos: a single horn pointed straight up at a concave dish, with small polystyrene beads or water droplets hovering in the gap between them at the pressure nodes.

The advantage of the single-axis standing-wave design is its simplicity: it requires only one emitter (or one coordinated cluster), a passive reflector, and a driving signal at a fixed frequency, with no need for complex phase control between multiple independent channels. The disadvantage is that it only really traps objects effectively along the central axis of the device, and the trapping strength in the lateral (sideways) direction is generally weaker than the trapping strength along the vertical axis, which can make single-axis levitators somewhat prone to objects drifting sideways out of the trap, particularly in the presence of air currents.

Modern multi-emitter standing-wave designs, of which TinyLev is the best-known example, improve substantially on this by replacing the single emitter and passive reflector with two opposing arrays of many small transducers, arranged so that their combined wavefronts form a much more tightly focused standing wave at the trapping location, with much stronger lateral confinement than a simple single-transducer design. Because each individual transducer in the array contributes only a small fraction of the total acoustic power, the array can achieve high total pressure at the focal point while keeping the voltage and power requirements for each individual transducer low, which is precisely why these designs can be driven at only around twenty volts rather than the hundred-plus volts needed by older single-transducer Langevin horn designs.

A second major architecture is far-field or single-sided levitation, in which there is no reflector at all, and instead an entire array of transducers on one side of the trapping region is driven with individually controlled phases so that their combined wavefronts constructively interfere at a chosen point in space, creating a localized high-pressure focus (or, for trapping purposes, a more complex field shape such as a pressure null surrounded by high pressure) some distance away from the array itself, without needing anything on the opposite side to bounce sound back. This is a considerably more sophisticated approach because it requires precise, independent control of the phase (and often amplitude) of every individual transducer element in the array, computed based on the desired field shape and the geometric position of each element relative to the target trapping point. It is, however, far more flexible: because there is no fixed reflector, the trapping location can be moved freely in three dimensions simply by recalculating the phase pattern sent to the array, and because there is no physical structure on the far side of the object, it becomes possible to work with an open architecture that leaves room for cameras, additional sensors, or human interaction with the levitated object.

Far-field, single-sided arrays are the foundation of what is now generally called holographic acoustic tweezing, a term borrowed directly from the analogous and earlier-developed field of holographic optical tweezers, where spatial light modulators reshape a laser beam to create multiple independently controllable optical traps. In the acoustic version, a two-dimensional array of individually addressable ultrasonic transducers — commonly built using standard 40 kHz elements arranged on a flat or slightly curved printed circuit board, each driven by its own phase-controlled channel — computes and emits the specific combination of phases needed to synthesize almost any desired pressure field shape in the volume of air in front of the array, including multiple simultaneous traps, traps that can be moved smoothly through space, and traps with more exotic shapes such as acoustic vortices (which carry orbital angular momentum and can be used to controllably rotate a levitated object) or acoustic bottle traps (which surround an object with a shell of high pressure on all sides at once, useful for trapping objects too large or dense for a simple standing-wave node).

A third and more specialized architecture, generally categorized as far-field levitation for large objects, departs from the small-particle assumption underlying the simplest Gor'kov analysis altogether and instead custom-tailors the acoustic field to the specific size and shape of a particular object larger than the wavelength, most often using multiple transducers arranged around the object (for example, in a tripod configuration providing forces from three different directions) or, in more recent and sophisticated implementations, using boundary-element numerical methods to calculate exactly what field shape, generated by a large phased array, will stably support an oddly shaped or unusually large object. This is currently one of the more computationally demanding but rapidly advancing branches of acoustic levitation research, and it is directly responsible for the largest and heaviest objects that have been successfully levitated to date, which are considerably larger than the simple wavelength-scale limit that governs basic standing-wave devices.

Each of these architectures represents a different point on a spectrum, trading off simplicity against flexibility. A hobbyist wanting to see levitation work with minimal cost and complexity will build a standing-wave design like TinyLev. A researcher wanting to move objects along programmed trajectories, split single traps into multiple independent ones, or manipulate the orientation of a levitated object will build a phased-array far-field design. And a researcher pushing the absolute size and mass limits of what can be levitated will use a large, custom-shaped, boundary-element-optimized field generated by a substantial multi-transducer array. Crucially, all three approaches rely on exactly the same underlying physics — acoustic radiation force arising from a carefully shaped, high-intensity sound field — and all three are achievable today using commercially available transducers, amplifiers, and control electronics, differing mainly in the sophistication of the control software and the number of independently addressable channels required.

The Hardware That Makes It Possible Today

If acoustic levitation has moved from an exotic laboratory technique to something buildable on a hobbyist budget, the reason is entirely a story about hardware becoming cheap, small, and controllable. It is worth walking through the actual components involved because the accessibility of each one is precisely what determines whether a working levitator is something a research group needs a grant to build, or something a motivated individual can assemble over a weekend.

The core active component in any acoustic levitator is the ultrasonic transducer itself, the device that converts an electrical signal into a mechanical sound wave. The overwhelming majority of modern designs use piezoelectric transducers, which exploit the piezoelectric effect: certain crystalline and ceramic materials, most commonly Lead Zirconate Titanate (PZT) ceramics, physically deform in response to an applied voltage, and this deformation, driven at high frequency, produces a sound wave. The transducers used in low-cost designs like TinyLev are the same mass-produced 40 kHz piezoelectric elements found in ultrasonic distance sensors, automotive parking-assist sensors, and ultrasonic cleaning baths — components manufactured by the millions for other industries and available from electronics suppliers for well under a dollar apiece when bought in the quantities needed for an array (commonly seventy-two to over one hundred units for a functional levitator). This is the single biggest factor separating modern acoustic levitation from its mid-twentieth-century predecessors: instead of needing custom, precision-machined, high-voltage Langevin horn transducers, a modern levitator can be built from parts originally intended for a car's backup sensor.

Driving these transducers requires a signal generation and amplification stage. Each transducer needs to be supplied with an alternating voltage at the ultrasonic operating frequency (typically 40 kHz, though other frequencies are used in specialized designs), and depending on the architecture, either all transducers can be driven in phase together (as in the simplest standing-wave designs) or each transducer needs its own independently controllable phase and sometimes amplitude (as in phased-array holographic designs). For a simple standing-wave levitator, this can be as basic as a single square-wave oscillator circuit, built around a low-cost microcontroller such as an Arduino, feeding a MOSFET-based driver stage that switches the transducers on and off at the target frequency, with a simple boost or step-up circuit to reach the roughly twenty-volt drive level needed for good acoustic output. This entire driver stage can be built for a small fraction of the cost of even a single component in an older, high-voltage Langevin-based system, and open-source circuit designs and firmware for exactly this purpose are freely published and widely replicated.

For more sophisticated phased-array designs, the electronics scale up considerably in complexity but remain built from broadly accessible components. Each channel in a phased array needs its own digitally controlled phase shift, which is typically implemented using a field-programmable gate array (FPGA) or a bank of digital signal processing chips that can generate many independently phased square-wave or sine-wave signals simultaneously, each routed through its own small amplifier stage to drive an individual transducer or small cluster of transducers. Systems built for academic research, such as the Ultraino open-source phased-array platform developed at the University of Bristol, have demonstrated that a functional 256-element phased array — capable of generating genuinely holographic, reconfigurable acoustic fields — can be built using open-source hardware designs, standard microcontroller boards, and commercially available transducers and driver chips, without requiring proprietary or exotic components. This kind of system represents a substantial step up in electronic complexity from a simple standing-wave levitator, involving printed circuit boards with dozens to hundreds of individually addressable output channels, but it remains squarely within the reach of a well-equipped university lab, a serious hobbyist electronics workshop, or a small engineering team, rather than requiring the kind of specialized national-laboratory infrastructure that some other precision physics techniques demand.

Beyond the transducers and driving electronics, the mechanical structure of a levitator matters more than one might expect. Because the trapping locations are determined by the precise geometry of the acoustic field, the physical frame holding the transducers in position needs to maintain that geometry with a precision on the order of a fraction of a millimetre, which is well within the capability of consumer-grade 3D printing, one of the technologies that has made the largest practical difference in making acoustic levitation buildable outside specialized machine shops. The original TinyLev design and its many derivatives and open-source variants rely entirely on 3D-printed transducer holders and frames, designed to hold each individual transducer at a precisely calculated angle so that the sound from every element in the array converges on the same focal region. Because 3D printers capable of the necessary precision are now inexpensive and widely available, this piece of the puzzle, which would once have required custom-machined metal or precision-molded plastic parts, has become essentially a non-issue for anyone with access to a basic desktop 3D printer or a local makerspace.

Control software is the final piece of the modern acoustic levitation stack, and it is here that the technology has perhaps advanced the furthest in the last several years. Early standing-wave levitators required no real-time computation at all — the field geometry was fixed by the physical arrangement of the transducers and reflector, and the only “control” involved was turning the driving signal on and off. Modern phased-array systems, by contrast, rely on software that computes, often in real time, the specific phase (and sometimes amplitude) that every individual transducer channel must be given in order to synthesize a desired field shape at a desired location, a computation that draws on exactly the same physical model — Gor'kov's potential and its extensions to multi-transducer geometries — described earlier in this article. Because this computation is, at its core, a matter of summing spherical wave contributions from many point sources with appropriate phase delays, it is well within the capability of a modern laptop or even an embedded microcontroller to compute in real time for arrays of up to several hundred elements, and open-source software libraries for exactly this purpose, developed by academic groups working on phased-array acoustic levitation, are freely available for researchers and hobbyists to build on rather than needing to derive the necessary wave-propagation mathematics from scratch.

Taken together, these four pieces — cheap mass-produced transducers, low-voltage driving electronics built from widely available components, 3D-printed mechanical structures, and open-source control software grounded in well-established acoustic theory — are the concrete answer to the question of how acoustic levitation could be possible today. Every one of these pieces already exists, is commercially available, and has been demonstrated to work in published, peer-reviewed, and independently replicated designs. The technology is not waiting on some undiscovered physical principle or unbuilt piece of exotic hardware. It is, in the most literal sense, already sitting on the shelves of electronics suppliers and hobbyist 3D-printing communities, waiting only to be assembled according to instructions that have already been written down.

Phased Arrays and Holographic Acoustic Tweezers

The most sophisticated and rapidly advancing branch of acoustic levitation technology today is the phased-array approach, and it deserves a closer, more detailed look because it represents the clearest path toward genuinely useful, programmable acoustic manipulation rather than simple, static hovering.

A phased array in this context is a flat or gently curved panel populated with dozens to hundreds of individually driven ultrasonic transducer elements, each of which can have its own phase (the timing of its oscillation relative to the other elements) and, in more advanced designs, its own amplitude, set independently under digital control. The physical principle at work is exactly the same beam-steering and beam-shaping mathematics used in fields as varied as radar, medical ultrasound imaging, and radio astronomy: by carefully choosing the relative timing of many small emitters, the combined wavefront they produce can be made to constructively interfere — reinforce itself — at essentially any chosen point in front of the array, while destructively interfering (cancelling out) elsewhere. Because the array can recompute these phase relationships electronically, essentially instantaneously and without moving any physical part, it can synthesize an enormous variety of acoustic field shapes on demand.

The landmark demonstration of this capability for airborne levitation was the 2019 publication by Asier Marzo and Bruce Drinkwater at the University of Bristol, describing a system they called holographic acoustic tweezers, built using two opposing 256-element phased arrays operating at 40 kHz. The system was able to independently trap and manipulate as many as twenty-five separate millimetre-scale particles simultaneously, moving each one along its own independently controlled trajectory, essentially treating each trapped particle as an individually addressable “voxel” of matter that could be positioned anywhere within the working volume between the two arrays. This capability directly mirrors what holographic optical tweezers achieved for microscopic particles roughly two decades earlier, but at a much larger physical scale, and with substantially higher trapping forces achievable per unit of input power because acoustic forces in air are simply much stronger than optical forces for objects of comparable size.

The practical significance of this shift from single-point, fixed-geometry levitation to fully programmable, multipoint holographic manipulation is considerable because it opens the door to applications that a simple standing-wave trap could never support. A holographic acoustic tweezer system can, in principle, be used to assemble small components into a designed three-dimensional structure by individually positioning each piece in sequence, held together temporarily or permanently once positioned, an approach that has been explored for microassembly and for volumetric, physically real 3D display concepts in which many small levitated particles are moved rapidly enough, and illuminated appropriately, to trace out a visible three-dimensional image in midair — a genuine, physical alternative to the purely projected images of conventional 3D displays.

Beyond simple point trapping, phased arrays can also generate more exotic and useful field topologies. An acoustic vortex, for example, is a field shape with a helical phase structure around a central axis, similar in concept to an optical vortex beam, and it produces a doughnut-shaped region of high pressure surrounding a central pressure null. Objects trapped in the central null of an acoustic vortex can be made to rotate by adjusting the vortex's rotational properties, giving researchers direct control over the orientation of a levitated object, not just its position — a capability with clear value for applications such as 3D printing or additive manufacturing of small components that need to be built up layer by layer while suspended, or for biomedical applications where controlling the orientation of a levitated biological sample matters for imaging or processing.

Another important field topology enabled by phased arrays is the acoustic bottle trap, in which the array generates a shell of high acoustic pressure completely surrounding a central low-pressure region on all sides, rather than relying on the planar pressure nodes of a simple standing wave. Because a bottle trap confines an object from every direction simultaneously rather than relying on a reflector on the opposite side, it is particularly well suited to single-sided array configurations where there is no reflecting surface at all, and it has been shown to be effective at trapping objects larger than would be stable in a conventional standing-wave node because the confinement comes from an enclosing shell of pressure rather than a single restoring gradient along one axis.

The control complexity of phased-array systems is nontrivial but, importantly, tractable with hardware that is now widely available. Driving a 256-element array requires 256 independently phase-controlled output channels, each capable of switching at 40 kHz with fine timing resolution, a requirement well within the capability of modern field-programmable gate arrays and even some microcontroller-based designs when the channel count is more modest. The computational load of calculating the required phase pattern for a desired field shape — effectively solving an inverse problem, working backward from “what field do we want” to “what phase must each element emit” — has been shown to be solvable in real time using standard optimization algorithms running on ordinary computing hardware, and open-source implementations of these algorithms, along with open hardware designs for the driver electronics themselves, have been published and are actively maintained by the academic groups that pioneered them, meaning a research team or sufficiently dedicated hobbyist does not need to reinvent this software and hardware stack from first principles in order to build a working holographic acoustic tweezer system today.

Overcoming the Wavelength Size Barrier

One of the most persistent misconceptions about acoustic levitation, even among people with some physics background, is that it is fundamentally limited to trapping objects that are tiny — dust-sized or smaller. This misconception is understandable given how the basic Gor'kov theory is usually taught, since it explicitly assumes the levitated particle is small compared to the acoustic wavelength, and for a 40 kHz system that wavelength is only about 8.6 millimetres. But this assumption is a simplification for mathematical tractability, not a hard physical law, and one of the more important research threads in acoustic levitation over the past decade has been systematically overcoming this apparent size limit.

The first major breakthrough came from recognizing that if a single transducer-and-reflector pair produces forces along essentially only one useful axis, adding additional transducers from different directions could supply the missing lateral stabilizing forces needed to keep a larger object from spinning out of the trap. A widely cited 2016 study demonstrated exactly this, using a tripod arrangement of three ultrasonic transducers to levitate a solid polystyrene sphere fifty millimetres in diameter — some 3.6 times larger than the acoustic wavelength being used — a size that would have been considered essentially impossible to stabilize using a conventional single-axis standing wave. The key insight was that a simple vertical standing wave alone can lift a large flat or spherical object against gravity, since the mechanism does not strictly require the object to be smaller than the wavelength, but it cannot supply the lateral restoring forces needed to keep the object centred and prevent it from toppling or drifting sideways; supplying those lateral forces from additional transducer directions solved the stability problem directly.

A second major line of research tackled the size barrier through what researchers termed an acoustic tractor beam, a phased single-emitter or emitter-array design capable of exerting a pulling, rather than purely holding, force on an object even at a distance greater than the acoustic wavelength. A widely reported 2018 study from the University of Bristol demonstrated that by generating a rapidly alternating acoustic vortex — described by the researchers as a tornado of sound, whose rotational direction is flipped back and forth fast enough to prevent that rotation from being transferred to the trapped object itself — it was possible to stabilize a levitated object considerably larger than had previously been achievable with tractor-beam-style designs, successfully and stably levitating a polystyrene sphere twenty millimetres across, roughly twice the acoustic wavelength at the 40 kHz operating frequency used, making it at the time the largest object ever held in a genuine acoustic tractor beam configuration. The researchers noted that the technique because it operated at an ultrasonic frequency safely above the range of human hearing, offered a plausible, safety-conscious pathway toward eventually extending similar techniques to hold and manipulate even larger objects.

More recent work has pushed further still by combining large-scale phased arrays with sophisticated numerical field-shaping techniques, most notably the boundary element method, which allows researchers to calculate precisely what acoustic field, generated by a large array of individually phased transducers, will produce a stable trap around an object of an arbitrary, specified size and shape — including irregular shapes, not just spheres or simple geometric forms. This computational approach represents a significant departure from the earlier, more heuristic tripod and tractor-beam designs because it treats the entire problem as a genuine inverse design exercise: given the desired object geometry and the physical layout of the available transducers, compute the field configuration that produces a stable equilibrium, rather than relying on intuition or trial-and-error transducer placement. Published results using this boundary-element approach have demonstrated stable levitation of expanded polystyrene objects with dimensions several times the acoustic wavelength, positioned tens of wavelengths away from the generating array, and including non-spherical shapes such as regular octahedra, substantially extending both the size and the geometric complexity of objects that can be acoustically trapped compared to earlier techniques.

It is worth being clear-eyed about what these advances do and do not mean for the near-term future of the technology. Levitating a two-centimetre polystyrene ball, while a genuine and impressive scientific achievement, is still a very long way from levitating anything with practical mass or density — a coin, a tool, a person. The underlying physics imposes a hard tradeoff: as an object's volume increases, its mass (and therefore the gravitational force that must be overcome) grows with the cube of its linear dimension for a fixed density, while the acoustic radiation force that can realistically be generated scales less favourably, especially once safety limits on sound intensity and the practical power-handling capacity of transducer arrays are taken into account. This is precisely why essentially all successful larger-object levitation demonstrations to date have used low-density materials, principally expanded polystyrene foam with a density around thirty kilograms per cubic meter, roughly thirty times less dense than water. Scaling these techniques to higher-density materials, let alone to objects with the mass of everyday solid objects, remains a substantial and unsolved engineering challenge, discussed further in the section on limitations later in this article. What has genuinely and durably changed, however, is the theoretical and engineering toolkit available to researchers: the wavelength is no longer treated as an unbreakable ceiling on object size, and a growing family of demonstrated multi-transducer, phased-array, and numerically optimized field-shaping techniques now exists to systematically push past it, in ways that were not available even fifteen years ago.

What Can Be Levitated Today

The theoretical discussion of forces and potentials only becomes meaningful once it is grounded in what has actually, physically been held aloft using sound, and the range of documented examples is broader than most people expect.

The most common and easiest category of object to levitate acoustically is small, low-density solids, of which expanded polystyrene foam beads are the standard laboratory workhorse, used in the overwhelming majority of published demonstrations and classroom videos precisely because their low density (a small fraction of water's density) and small typical size make them easy to trap with modest, low-power hardware. Similarly, easy are small pieces of many lightweight organic and inorganic materials: grains of rice, small paper fragments, seeds, and similarly light, millimetre-scale objects have all been routinely levitated in demonstration and research settings.

Liquids represent a particularly important and heavily studied category because acoustic levitation offers something no ordinary container can: the ability to hold a liquid droplet in place with absolutely no contact with any solid surface. This matters scientifically because a liquid droplet in contact with a container wall behaves differently than a truly free droplet — the wall can act as a nucleation site that triggers unwanted crystallization, can contaminate a chemically sensitive sample, or can simply distort the droplet's natural spherical shape through surface adhesion. Acoustically levitated water droplets, and droplets of a wide range of other liquids including various solvents, biological fluids, and pharmaceutical formulations, have been used extensively to study evaporation dynamics, to observe super cooling (keeping a liquid below its normal freezing point without ice crystals forming, which is much easier to sustain in a container-free environment), and to watch the crystallization process of dissolved solutes as a droplet slowly dries, since the resulting particle forms without any container-induced asymmetry.

Small biological specimens form another well-documented category. Insects, particularly small ones such as ants and small beetles, have been successfully and repeatedly levitated in published research and demonstration videos, generally shown to survive the experience without apparent harm, since the acoustic pressures involved, while intense enough to produce a significant net force on a lightweight object, do not correspond to physically damaging levels of energy deposition in living tissue at the exposure durations typically used. More scientifically significant has been the acoustic levitation of biological cell cultures, small fish embryos such as zebrafish, and even the demonstrated ability to hold small organisms in a stable, contact-free environment for extended observation periods, an application directly relevant to research on how organisms develop and behave in reduced-gravity or contact-free conditions, discussed further in the section on space and microgravity research below.

Electronic and manufacturing-relevant objects have also been demonstrated as levitation targets, an application area with obvious practical relevance to electronics assembly. Small electronic components, printed circuit board elements, and similarly sized manufactured parts have been shown to be levitatable and manipulable using phased-array systems, pointing toward a future in which delicate components could be positioned, oriented, and assembled in a completely contact-free manner, eliminating the risk of mechanical damage, electrostatic discharge from handling equipment, or contamination that can occur with conventional pick-and-place robotic assembly tools.

At the upper end of the size and mass spectrum, the case studies discussed in the previous section — the fifty-millimetre tripod-levitated polystyrene sphere, the twenty-millimetre acoustic-tractor-beam-levitated sphere, and the more recent boundary-element-optimized levitation of centimetre-scale spheres and octahedra several wavelengths from their generating array — represent the current frontier of what has actually been achieved, and they remain, without exception, objects made of low-density expanded polystyrene foam rather than any denser, more practically useful material.

It is worth noting explicitly what has not been achieved, both to keep expectations grounded and to accurately map the boundary of current capability. No published, peer-reviewed acoustic levitation demonstration has successfully and stably levitated an object with the density of a typical solid metal, wood, or even most plastics at any size larger than roughly a few millimetres because the gravitational force such an object experiences at any practical size quickly outpaces the force achievable with currently demonstrated acoustic field intensities. No acoustic levitation system has come remotely close to levitating anything approaching the mass of a human being or even a small household object of ordinary density, and claims occasionally seen in popular media suggesting imminent “human levitation” substantially overstate what the underlying physics and current engineering realistically support, a point discussed further in the limitations section below. What has been achieved, and achieved robustly and repeatably across many independent research groups, is reliable, contact-free suspension and manipulation of a wide range of small, generally low-to-moderate-density objects — droplets, foam spheres, small organisms, small electronic components, and biological samples — using hardware that, as described earlier, is genuinely buildable today from commercially available parts.

Building an Acoustic Levitator Today

Perhaps the most direct way to answer “how could acoustic levitation be possible today” is to walk through, concretely, what building a basic working levitator actually involves, using the well-documented and widely replicated TinyLev-style design as the reference point, since it remains the most accessible, best-tested, and most thoroughly published open-source approach available.

The core of the design is a pair of opposing concave arrays of small 40 kHz ultrasonic transducers, commonly using around thirty-six to forty transducers per array (roughly seventy-two total), each individually inexpensive and available from general electronics suppliers, of the type manufactured for ultrasonic distance-sensing applications. The transducers are mounted in two 3D-printed holders shaped as shallow concave bowls so that the sound emitted by every individual transducer in the array is aimed, geometrically, toward a common focal point in the space between the two arrays — this geometric focusing is what allows dozens of cheap, individually modest transducers to combine into a strong, tightly localized standing wave, rather than a diffuse, weak field. The two arrays are mounted facing each other, separated by a gap tuned to be a whole number of half-wavelengths at the 40 kHz operating frequency (a spacing of roughly four to five centimetres is typical), which establishes the standing-wave condition needed for stable trapping.

Driving the transducers requires a modest electronics stage: since all the transducers in this simple design are driven in phase together (there is no need for the individually addressable phase control required by a more advanced holographic phased array), a single square-wave signal at the 40 kHz operating frequency, generated by an inexpensive microcontroller such as an Arduino, can be amplified through a simple MOSFET half-bridge driver stage and boosted to a drive voltage of roughly twenty volts, well, within the range of a small, inexpensive DC power supply. This entire electronics package — microcontroller, driver circuit, and power supply — can be assembled using widely available components and published open-source circuit schematics, without requiring specialized fabrication capability beyond basic soldering and simple circuit board assembly, which is achievable by an individual with modest electronics hobbyist experience or, more simply still, by ordering a pre-assembled printed circuit board from a commercial fabrication service using the openly published design files.

Assembly, once the parts are in hand, primarily involves 3D printing the two transducer-holder shells (design files for these are published openly and can be printed on essentially any consumer-grade fused-deposition 3D printer), soldering the transducers into their mounting positions with the correct wiring to ensure they are all driven with the same polarity and phase, wiring the two arrays to the driver circuit, and mechanically fixing the two arrays at the correct separation distance using a simple frame, which can itself be 3D printed or assembled from basic hardware-store materials such as threaded rod and standoffs. Total build time for someone with basic soldering and 3D-printing experience, following the published documentation, is commonly reported in the range of a few hours to a weekend, and the total component cost for a basic build has been documented at a level comparable to a modest consumer electronics purchase — considerably less than the cost of the specialized, high-voltage Angevin-transducer-based levitators that would have been the only option a generation earlier. More recent open-source variants, such as the AcoustiFly design developed specifically for educational use, have pushed the cost even lower, to roughly the price of a fast-food meal, by simplifying the design to a single transducer with a passive reflector rather than a full multi-transducer array, trading some lifting capability for even greater simplicity and cost reduction, explicitly aimed at making the phenomenon accessible for classroom demonstrations and individual students rather than requiring laboratory-grade equipment.

Once assembled and powered on, a TinyLev-style device requires no manual tuning or calibration in the way older high-voltage designs did; the low-cost transducers used are sufficiently well-matched in their manufactured resonant frequency, and the drive circuit sufficiently simple, that the device begins producing a usable standing wave essentially as soon as it is switched on. Objects are introduced into the trap using tweezers or a small dropper (for liquids), positioned near one of the visible or calculated pressure-node planes between the two transducer arrays, where they will settle into stable levitation once released, held in place by the restoring force described earlier in this article. Published results from the original TinyLev research demonstrated stable levitation of water droplets, small fused-silica glass spheres, small insects, and small electronic components, with some samples maintained in stable levitation for periods exceeding two hours without any active intervention or feedback control, a testament to the genuine stability of the underlying trapping potential once properly established.

For those wanting to go beyond simple fixed-point levitation and explore the programmable, holographic capabilities described earlier in this article, the natural next step is a phased-array build, which is substantially more involved but still grounded entirely in openly published, replicable designs. The Ultraino platform, an open-source phased-array system developed specifically for this purpose by the same University of Bristol research group behind TinyLev, provides published hardware schematics and firmware for driving a two-dimensional array of individually phase-controlled 40 kHz transducers, along with accompanying software for computing the phase patterns needed to synthesize a desired trapping field, whether a single movable trap, multiple simultaneous traps, or one of the more exotic field topologies such as acoustic vortices or bottle traps described earlier. Building a system at this level of sophistication requires meaningfully more electronics assembly work — populating a printed circuit board with many individually addressable driver channels — but it remains, crucially, built entirely from openly documented designs and commercially available components, rather than requiring proprietary technology or specialized fabrication capability unavailable outside a small number of institutions.

The overall picture that emerges from actually walking through a build is one of remarkable, and quite recent, accessibility. As little as fifteen years ago, replicating even the simplest standing-wave acoustic levitation demonstration would have required access to a specialized acoustics or ultrasonics laboratory, custom-machined high-voltage Langevin transducers, and considerable expertise in mechanical tuning. Today, the same basic phenomenon, and increasingly the far more capable programmable phased-array version of it, can be built by a single motivated individual with a 3D printer, basic soldering skills, and a modest budget for electronic components, following freely available, peer-reviewed, and independently replicated open-source designs. This is the concrete, hardware-level answer to how acoustic levitation could be possible today: not through some undiscovered principle, but through the democratization of exactly the components needed to apply well-established physics.

Real-World and Emerging Applications

Acoustic levitation is not being pursued purely as a demonstration of interesting physics; it is being actively developed and, in several cases, already deployed as a practical tool across a genuinely diverse range of fields, precisely because the contact-free, containerless manipulation it offers solves real problems that conventional handling methods cannot.

Pharmaceutical research and formulation science represent one of the most mature application areas. Drug formulation frequently requires studying how a compound behaves as a droplet dries, crystallizes, or is processed into its final solid form, and doing this in a container introduces exactly the kind of wall-contact artifacts described earlier — nucleation sites that trigger premature or unnatural crystallization, contamination from container surfaces, and loss of sample to adhesion on container walls, which matters enormously when working with expensive or scarce compounds available only in microgram quantities during early-stage drug development. Acoustically levitating a single droplet of a drug formulation allows researchers to observe its natural drying and crystallization behaviour in complete isolation from any surface, using techniques such as Raman spectroscopy or other optical probing methods trained directly on the levitated droplet, and this approach has been used in published pharmaceutical research to study polymorphism (the tendency of a compound to crystallize into different structural forms with different pharmacological properties), a property that matters enormously for drug efficacy, stability, and manufacturability.

Materials science continues to rely on acoustic levitation for containerless processing, extending the tradition established in the 1970s and 1980s described earlier in this article's historical section, but now with far more accessible and controllable equipment. Studying the solidification behaviour of metal alloys, the formation of amorphous (glassy) solid phases that require rapid, undisturbed cooling from a melt, or the fundamental thermophysical properties of materials at extreme temperatures, all benefit from a processing environment with zero contact between the sample and any container wall, since contact points are almost always where undesired crystal nucleation begins. Modern phased-array and multi-transducer techniques, capable of levitating larger, more strongly heated samples than earlier single-axis designs, are extending this application to a wider range of materials and sample sizes than was previously practical.

Biomedical research and cell biology form a rapidly growing application area, particularly as phased-array holographic acoustic tweezer systems mature. The ability to trap, position, and move individual cells, cell clusters, or small tissue samples without any physical contact offers clear advantages over traditional mechanical handling tools such as pipettes or micromanipulators, which can damage delicate biological material or introduce contamination. Published research has explored using acoustic tweezers for applications including the patterning of cells into designed arrangements for tissue engineering, holding microorganisms in a fixed position for extended microscopic study without the stress of mechanical restraint, and — in one particularly forward-looking line of research — using single-beam acoustic tweezers operating through biological tissue to manipulate nanoparticle drug carriers for targeted drug delivery, an approach that, unlike magnetic drug-targeting techniques, does not interfere with magnetic resonance imaging and could in principle be used to guide therapeutic payloads to specific locations within the body using externally applied, precisely focused ultrasound fields.

Three-dimensional bioprinting is a specific and increasingly discussed extension of this biomedical work, in which acoustic tweezers are used not merely to hold cells in place but to actively position individual cells or small cell clusters into a designed three-dimensional arrangement, effectively “printing” a tissue structure by acoustically placing living material into a scaffold or matrix layer by layer, an approach that avoids the mechanical stress that conventional extrusion-based bioprinting nozzles can impose on delicate cells, and that can, in principle, achieve finer spatial control over individual cell placement than extrusion-based methods allow.

Electronics manufacturing and precision assembly represent another practically significant frontier. As electronic components continue to shrink and become more delicate, and as manufacturing processes increasingly involve materials sensitive to mechanical stress, heat, or electrostatic discharge from conventional robotic handling tools, contact-free acoustic manipulation offers a genuinely different approach to component handling: components can be picked up, moved, rotated, and placed using programmable acoustic fields generated by a phased array, without ever touching a mechanical gripper, vacuum nozzle, or other physical handling tool. This capability has been explored directly in published phased-array research demonstrating the manipulation of small electronic components using holographic acoustic tweezer systems, and it points toward a longer-term manufacturing application in precision electronics assembly, particularly for components too small, delicate, or electrostatically sensitive for conventional pick-and-place robotics.

Volumetric and physically real three-dimensional displays represent one of the more visually striking and widely publicized applications of phased-array acoustic levitation. By using a phased array to rapidly move one or several small trapped particles through a three-dimensional volume of space fast enough that the human eye perceives a continuous, glowing trace rather than a single moving point — an approach conceptually related to persistence-of-vision displays — and illuminating the moving particle with coloured light synchronized to its position, researchers have demonstrated genuinely three-dimensional images formed from physically real, moving matter suspended in midair, viewable from any angle without special glasses or a projection screen, a fundamentally different approach to 3D display technology than the illusory depth cues used in conventional stereoscopic or autostereoscopic displays. While still a research demonstration rather than a commercial product, this application illustrates the kind of genuinely novel capability that becomes possible once acoustic levitation moves from single, fixed-point trapping to fully programmable, multipoint holographic manipulation.

Food science and consumer applications, while less prominently developed than the areas above, have also been explored, including the use of acoustic levitation to study the drying and preservation behaviour of food-relevant liquid droplets, and speculative but seriously discussed concepts for contact-free food handling in specialized manufacturing or hygiene-sensitive contexts, where avoiding any physical contact between a handling tool and a food product carries clear sanitary advantages.

Across all of these application areas, a consistent theme emerges: acoustic levitation is valuable not because it can lift heavy objects — it fundamentally cannot, for the physical reasons discussed throughout this article — but because it offers something genuinely difficult to achieve any other way: precise, programmable, completely contact-free manipulation of small to modest-sized objects, liquids, and biological material, in a way that avoids contamination, mechanical stress, and the container-wall artifacts that plague conventional handling techniques. This is where the real, near-term practical value of the technology lies, and it is exactly the kind of application space where the accessible, low-cost hardware described earlier in this article is already being put to genuine scientific and, increasingly, industrial use.

Acoustic Levitation in Space and Microgravity Research

One of the more scientifically important and conceptually interesting applications of acoustic levitation connects directly back to gravitational physics: using acoustic trapping as a tool to study, simulate, or support research into microgravity and reduced-gravity environments, both aboard spacecraft and, increasingly, in ground-based laboratories seeking a cheaper and more accessible alternative to actual spaceflight.

The historical roots of this application, described earlier in the historical section of this article, trace back to Taylor Wang's acoustic containment experiments flown aboard the Space Shuttle in the 1980s and continued through the United States Microgravity Laboratory missions in the 1990s, which used acoustic levitation specifically because it offered a containerless environment for studying fluid dynamics in an environment where gravity was no longer distorting the results — a use case where acoustic levitation was not simulating microgravity but rather being used as a complementary containment tool within a genuine microgravity environment, avoiding the contact and contamination issues that a physical container would introduce even in orbit.

A more recent and, in some ways, more broadly useful line of research inverts this relationship, using ground-based acoustic levitation as a tool to simulate certain aspects of a microgravity or reduced-gravity environment without ever leaving Earth. While acoustic levitation does not eliminate gravity in any real sense — the levitated object's weight is still being actively counteracted by the acoustic radiation force, moment to moment, rather than gravity itself being switched off — the containerless, contact-free, mechanically unconstrained environment it creates shares important practical similarities with a genuine microgravity environment for certain kinds of biological and fluid studies, particularly ones concerned with the absence of mechanical stress, contact-induced stimulation, or sedimentation effects that would otherwise be present for an object resting on a solid surface under normal gravity.

Published research reviewed under this framing has explored acoustically levitating small biological specimens, including cell cultures and small aquatic organisms, and observing developmental or behavioural changes that parallel, at least in some respects, changes observed in genuine spaceflight microgravity experiments, positioning ground-based acoustic levitation as a considerably cheaper, more accessible, and more readily repeatable screening tool for identifying biological effects worth further investigation in genuine microgravity conditions aboard the International Space Station or other spacecraft, rather than as a complete substitute for actual spaceflight research. This distinction matters and is worth stating clearly: acoustic levitation is a tool for studying certain contact-free and unsedimented aspects of biological and fluid behaviour, and a genuinely useful complement to spaceflight research, but it is not literally gravity manipulation, and objects levitated acoustically on Earth's surface are still experiencing the same one-gravity gravitational acceleration as everything else in the room — they are simply being held aloft against it by a separate, actively applied force, in exactly the same physical sense that a helium balloon or a thrown ball briefly resists gravity, just through a different and considerably more precisely controllable mechanism.

Beyond biological research, acoustic levitation continues to be discussed and explored as a genuinely useful in-space manufacturing and materials-handling tool for future spacecraft and orbital platforms, extending its historical containerless-processing role. In an actual microgravity environment, where objects do not need to be lifted against gravity but merely held in a fixed, stable position and prevented from drifting, the power requirements for acoustic manipulation are considerably lower than they are for Earth-based levitation against full gravitational weight, making the technique potentially even more practical for orbital applications such as containerless materials processing, precision assembly of small components, or contact-free handling of biological or chemical samples aboard a space station or spacecraft, than it already is on the ground.

Challenges, Limitations, and Safety Considerations

No honest treatment of how acoustic levitation could be possible today would be complete without a clear-eyed accounting of what it cannot yet do, and what genuine engineering and physical obstacles stand in the way of scaling the technology beyond its current, well-demonstrated capabilities.

The most fundamental limitation, discussed at several points throughout this article, is the mass and density ceiling imposed by the physics of acoustic radiation force. Because the achievable acoustic pressure in air is bounded by practical limits on transducer power, array size, and — critically — safety considerations around sound intensity, and because the gravitational force on an object grows with its volume while the maximum achievable acoustic force does not grow proportionally without a corresponding increase in field intensity and array size, there is a hard, physics-based ceiling on how large or dense an object can be levitated with any given piece of hardware. This is why, as detailed earlier, essentially every demonstration of larger-object acoustic levitation has relied on low-density expanded polystyrene foam rather than denser, more practically useful materials, and why claims of imminent human-scale acoustic levitation substantially overstate current capability; scaling to human mass and density would require acoustic intensities and array sizes far beyond anything demonstrated or currently practical, and would run directly into serious safety limits on sound exposure well before the required lifting force could plausibly be reached.

Air currents and environmental disturbance represent a more mundane but practically significant limitation, particularly for the lower-power, simpler standing-wave designs most accessible to hobbyists and educators. Because the trapping force at the edge of a stable region is finite and the restoring force weakens as an object is pushed toward the boundary of its potential well, even modest air currents — from an open window, an air conditioning vent, or simply someone walking near the device — can be enough to dislodge a lightly trapped object from a weaker levitator, which is one of the reasons more sophisticated designs favor stronger, more tightly focused multi-transducer arrays over simple single-transducer setups, and why serious laboratory demonstrations are typically conducted in still-air conditions or enclosed housings.

Heat management is a genuine engineering constraint, particularly for higher-power systems attempting to levitate larger or denser objects. Piezoelectric transducers are not perfectly efficient, and a meaningful fraction of the electrical energy driving them is dissipated as heat rather than converted into acoustic output; as arrays are driven harder to achieve the higher intensities needed for larger or heavier levitated objects, transducer heating becomes a limiting factor both because excessive heat can damage or degrade the transducers themselves and because, as noted in the historical discussion of older Langevin-horn designs, heating shifts a transducer's resonant frequency, which can detune the standing wave and destabilize the trap if not actively compensated for.

Acoustic streaming presents a more subtle physical limitation that becomes increasingly important at higher sound intensities. Beyond a certain intensity threshold, a sound field in air begins to induce actual bulk fluid motion — a steady flow of air driven by the sound field itself, distinct from the oscillatory back-and-forth motion of the sound wave — and this streaming flow can destabilize a levitated object, particularly a liquid droplet, by introducing an asymmetric, disruptive force that the simple standing-wave trapping model does not account for. Managing acoustic streaming effects becomes an increasingly important design consideration as researchers push toward higher acoustic intensities to levitate larger or denser objects, and it represents one of the genuine physical (rather than purely engineering) obstacles to indefinitely scaling up acoustic levitation power.

Safety considerations around sound exposure deserve direct and serious treatment, since they are sometimes glossed over in popular discussions of the technology. The high-intensity ultrasonic fields used in acoustic levitation, while inaudible to humans at the standard 40 kHz operating frequency used in most designs, are not necessarily harmless simply because they cannot be heard; sustained exposure to very high-intensity ultrasound has been associated with effects including localized heating of tissue and, at sufficiently high intensities, potential harm to hearing-adjacent structures or other biological effects, and researchers working with high-power acoustic levitation equipment, particularly larger phased-array systems capable of generating the higher intensities needed for larger levitated objects, take deliberate precautions around exposure limits and equipment shielding. This safety consideration becomes considerably more acute for any research direction exploring lower-frequency, audible-range acoustic levitation (used in some large-object demonstrations specifically because lower frequencies correspond to longer wavelengths, which helps address the size-limit challenges discussed earlier), since audible-range high-intensity sound carries well-established and much more thoroughly documented risks to human hearing, meaning that any serious future push toward larger-scale acoustic levitation using lower frequencies will need to grapple directly and carefully with human safety exposure limits, not merely transducer engineering.

Finally, control and stability challenges scale up considerably as system sophistication increases. While a simple, fixed-geometry standing-wave levitator like TinyLev requires essentially no active control once assembled, the phased-array and holographic systems capable of the more advanced manipulation described earlier in this article require real-time computation, precise multichannel phase synchronization, and, for applications involving moving traps or multiple simultaneously manipulated objects, feedback systems to track object positions and correct for drift or disturbance, all of which represent genuine, nontrivial control-engineering challenges that remain active areas of ongoing research rather than fully solved problems, even as the underlying hardware components required to build such systems have become broadly accessible.

Comparing Acoustic Levitation with Other Levitation Methods

Acoustic levitation is one of several distinct physical techniques capable of suspending objects without mechanical support, and understanding how it compares to its main alternatives — magnetic levitation, electrostatic levitation, aerodynamic levitation, and optical levitation — helps clarify exactly what niche it fills and why it has become the technique of choice for the specific applications described in this article, rather than a universal replacement for all forms of contact-free manipulation.

Magnetic levitation, familiar from maglev trains and a range of industrial applications, works by using magnetic fields to counteract gravity on an object with appropriate magnetic properties, either through simple attraction or repulsion between permanent magnets and electromagnets, or through more sophisticated diamagnetic or superconducting levitation techniques. Magnetic levitation can support considerably larger masses than acoustic levitation — entire train cars, in the case of maglev transportation — but it is fundamentally restricted to materials with the right magnetic properties: ferromagnetic materials for simple attractive/repulsive levitation, or materials with sufficiently strong diamagnetic response (such as water, and by extension living tissue, which is how the well-known frog-levitation demonstrations using extremely powerful electromagnets work) for diamagnetic levitation. Acoustic levitation's key advantage over magnetic levitation is material generality: it works, in principle, on any object regardless of its electrical or magnetic properties, as long as it presents enough surface area relative to its mass for the achievable acoustic pressure to overcome gravity, making it useful for manipulating non-magnetic materials such as most plastics, glass, biological tissue, and liquids that magnetic levitation simply cannot address without extraordinarily strong and often impractical field strengths.

Electrostatic levitation uses electric fields to support a charged object against gravity, and has been used effectively in specialized materials-science and containerless-processing contexts, particularly for levitating molten metal samples in vacuum or controlled-atmosphere environments where the sample can be reliably charged and where there is no air to interfere with the electric field. Electrostatic levitation can achieve impressive results for specific, well-controlled sample types, particularly in vacuum, but it generally requires the levitated object to hold a controlled electric charge, which is not always straightforward to establish and maintain, particularly for insulating or biological materials, and the technique is less naturally suited to open-air, ambient-condition operation than acoustic levitation, which requires no special atmosphere and no charging of the sample at all.

Aerodynamic levitation, which suspends an object on a carefully controlled jet of gas, is mechanically simple and has been widely used in materials science for containerless processing of high-temperature melts, since the gas jet can be maintained even at very high sample temperatures where other levitation methods might struggle. Its principal limitation compared to acoustic levitation is control precision and stability: a simple gas jet provides comparatively coarse positional control and can be prone to sample rotation or oscillation instabilities that require careful nozzle design and gas-flow tuning to manage, whereas acoustic levitation, particularly in its phased-array form, offers far more precise, programmable, three-dimensional positional control.

Optical levitation, using the radiation pressure of focused laser light (the same physical principle underlying optical tweezers), is the technique most philosophically similar to acoustic levitation, since both rely on radiation pressure from a wave phenomenon rather than magnetic, electrostatic, or aerodynamic forces. The essential practical difference, discussed earlier in this article's physics section, is one of achievable force per unit of input power: because acoustic waves in air can carry far higher effective radiation pressure at reasonable power levels than laser light can in the same medium, optical levitation is generally restricted to genuinely microscopic particles — single cells, small beads a few micrometers across, individual molecules in some specialized configurations — while acoustic levitation comfortably handles objects from the submillimetre scale up to multiple centimetres, a difference of several orders of magnitude in the practical size range each technique addresses. Optical levitation retains distinct advantages for applications requiring extraordinarily fine force control at the molecular or single-cell scale, such as measuring the mechanical properties of individual DNA molecules or motor proteins, applications where acoustic levitation's comparatively larger minimum practical trap size and lower force resolution make it less suitable.

Taken together, this comparison clarifies acoustic levitation's genuine niche: it is the levitation technique of choice specifically for contact-free manipulation of millimetre-to-centimetre-scale objects and liquid samples, made of essentially any material, in an ordinary open-air, ambient-temperature, ambient-pressure environment, without requiring the sample to be magnetic, electrically charged, or restricted to the genuinely microscopic scale that optical methods are best suited to. This is a real and valuable niche, occupied by acoustic levitation more effectively than any competing technique, but it is also a bounded one, and understanding those bounds is essential to correctly assessing what the technology can and cannot practically deliver today.

Multi-Object Manipulation and Recent Breakthroughs

The field has continued to advance meaningfully in the very recent past, and it is worth closing the technical portion of this article with a look at where the frontier currently stands, since some of the most significant conceptual advances have come only in the last year or two.

A particularly notable recent breakthrough, reported by researchers at the Institute of Science and Technology Austria in late 2025, addressed a long-standing practical limitation of multi-object acoustic manipulation: the tendency of multiple small objects levitated simultaneously within the same acoustic field to drift together and clump, driven by secondary acoustic forces (sometimes called Bjerknes forces) that arise from the way nearby levitated objects scatter and re-radiate sound, subtly attracting each other in ways that a simple single-particle trapping analysis, such as the basic Gor'kov formulation described earlier in this article, does not capture. Because these inter-particle attraction effects become significant whenever multiple objects are trapped close together, they had previously placed a practical limit on how many separate objects could be simultaneously and independently manipulated in the same acoustic workspace before they began interfering with each other's stability. The Austrian team's approach introduced controlled electrical charging of the levitated particles as an additional tool, using electrostatic repulsion between the individually charged objects to counteract the acoustic clumping tendency and maintain separation between multiple simultaneously levitated objects, effectively combining two different physical principles — acoustic radiation trapping and electrostatic repulsion — within a single system to solve a problem that acoustic forces alone could not adequately address. The research team explicitly framed this advance as significant for fields such as microrobotics and materials science that depend on being able to reliably position multiple small building blocks relative to one another without them sticking together, a capability directly relevant to the assembly and 3D-printing-style applications discussed earlier in this article.

This kind of hybrid approach — combining acoustic levitation with a second physical principle to overcome a limitation specific to sound-based trapping alone — reflects a broader and increasingly important trend in the field: rather than treating acoustic levitation as a self-contained, standalone technology, current research increasingly treats it as one tool within a larger combined toolkit, paired where useful with electrostatic, optical, or even magnetic techniques to address specific limitations that pure acoustic trapping cannot solve on its own. This trend is likely to continue and deepen as the technology matures, since it allows researchers to draw on the genuine strengths of acoustic levitation — strong forces per unit power, material generality, and increasingly sophisticated programmable field shaping — while compensating for its specific weaknesses, such as inter-particle clumping in dense multi-object configurations, using complementary physical mechanisms.

Parallel progress continues on the boundary-element and numerical field-optimization techniques described earlier in the discussion of overcoming the wavelength size barrier, with ongoing published research extending the range of object sizes, shapes, and densities that can be stably levitated using increasingly large and sophisticated phased-array hardware, alongside continuing refinement of the underlying computational methods used to design the required fields. The overall trajectory across essentially every major research group active in this space points in a consistent direction: larger effective working volumes, more independently controllable simultaneous traps, greater geometric flexibility in the shapes and sizes of objects that can be handled, and increasingly sophisticated hybrid approaches that combine acoustic trapping with complementary physical techniques where pure acoustics falls short.

Realistic Future Prospects

Looking forward from today's demonstrated capabilities, several development directions appear both scientifically plausible and grounded in trends already visible in current published research, rather than purely speculative extrapolation.

Continued scaling of phased-array systems, both in the number of individually controllable transducer channels and in the overall physical size of the array, is a near-certain direction of continued development, following the same broader technology trend that has driven down the cost and increased the capability of digital signal processing and multichannel electronic control systems across many unrelated industries. As array channel counts continue to grow and the associated control electronics continue to shrink in cost, the practical working volume, maximum number of simultaneously manipulated independent objects, and achievable field complexity of holographic acoustic tweezer systems are all likely to continue increasing, extending the kind of multi-object, programmable manipulation capability demonstrated by systems such as the Bristol holographic acoustic tweezers described earlier in this article.

Continued integration of acoustic levitation with complementary physical techniques, following the pattern of the recent electrostatic-charging breakthrough described in the previous section, represents another highly plausible and already-emerging direction, likely to be extended to other combined approaches — potentially including magnetic assistance for ferromagnetic components, optical techniques for the very smallest particles at the boundary between the acoustic and optical size regimes, or active feedback control systems using machine vision to track and correct the position of levitated objects in real time, compensating for the air-current and disturbance sensitivity discussed in the limitations section, rather than relying purely on the passive stability of the trapping potential.

Continued growth in biomedical and pharmaceutical applications appears highly likely given the substantial and growing body of published research already summarized earlier in this article, particularly as holographic acoustic tweezer systems continue to mature toward the kind of precise, programmable, multi-object manipulation needed for genuinely useful applications such as targeted 3D bioprinting of tissue structures or in-body targeted drug delivery using focused ultrasound, both of which remain active, serious research directions rather than settled, deployed technologies, but ones with a clear and demonstrated developmental trajectory rather than a purely speculative one.

Realistic limits should also be stated plainly as part of any honest forward-looking assessment. The fundamental physical tradeoff between object mass and achievable acoustic force, discussed at length in the limitations section, is not going to be dissolved by incremental engineering improvement; levitating genuinely dense, human-scale objects with acoustic forces alone would require sound intensities well beyond what is safe for humans to be near, and there is no currently plausible engineering pathway around this fundamental constraint using sound alone. Popular-media framings that suggest acoustic levitation is on an inevitable path toward science-fiction-style human levitation should be treated with appropriate skepticism; the genuine and exciting trajectory of the field lies in extending precise, programmable, contact-free manipulation of small-to-modest objects, liquids, and biological material — a genuinely valuable and rapidly maturing capability in its own right — rather than in overcoming gravity for large, dense, everyday objects, a goal the underlying physics does not currently support and shows no clear sign of supporting in the foreseeable future.

Design Choices That Determine Success

Beyond the broad architectural categories already discussed, a handful of more granular design decisions determine whether a real, physically assembled acoustic levitator performs well or performs poorly, and these details matter enough to anyone actually building a system that they deserve dedicated treatment.

Frequency choice is the first and most consequential decision, and while 40 kHz has become something close to a de facto standard for airborne acoustic levitation, it is worth understanding why, since the reasoning illuminates a genuine set of engineering tradeoffs rather than an arbitrary convention. Lower frequencies produce longer wavelengths, and because the maximum practical size of a stably trapped object scales with wavelength (even accounting for the size-barrier-breaking techniques discussed earlier, which still benefit from a larger baseline wavelength), lower frequencies are attractive for anyone wanting to levitate larger objects. But lower frequencies below roughly 20 kHz fall within the range of human hearing, and sustained exposure to a loud audible tone is not merely an annoyance but a genuine occupational hazard, meaning any levitator operating at audible frequencies and high intensity needs to be treated with the same seriousness as any other loud industrial noise source, including hearing protection for anyone nearby and careful attention to permissible exposure durations. Higher frequencies, conversely, shrink the wavelength and correspondingly shrink the maximum practical object size, but they offer two significant compensating advantages: they are inaudible to humans even at high intensity, removing the hearing-safety concern that dogs lower-frequency designs, and they allow the transducers themselves to be physically smaller, since a piezoelectric transducer's physical size is generally linked to the wavelength it is designed to emit efficiently, meaning higher-frequency arrays can pack more individual emitters into a given panel area, increasing the achievable total acoustic power and field complexity for a fixed array footprint. The 40 kHz convention represents a practical sweet spot within this tradeoff space: safely above the audible range for the vast majority of adult listeners, low enough to still produce a manageable millimetre-scale wavelength suitable for trapping small-to-moderate objects, and matched to a mature, mass-produced, low-cost transducer ecosystem originally developed for ultrasonic ranging applications, which is itself a significant practical advantage independent of the underlying acoustic physics.

Reflector and array curvature is a second design detail that matters more than casual descriptions of “a transducer and a reflector” might suggest. A flat reflector or a flat array of transducers produces a standing wave, but the resulting field, while functional, is not optimally concentrated: acoustic energy spreads somewhat as it propagates, and a flat source does not counteract this spreading. Curving the reflector concavely, or arranging an array of individual transducers on a concave shell so that each one is aimed at a common focal point, counteracts this natural spreading and concentrates acoustic energy much more tightly at the intended trapping location, substantially increasing the achievable pressure amplitude, and therefore the achievable trapping force, for a given total input power. This is precisely why the TinyLev design and its many derivatives use concave transducer-holder shells rather than flat panels, and it is a design choice validated repeatedly across published research: for a fixed number of transducers and a fixed total power budget, a well-curved, properly focused array will out-perform a flat one by a substantial margin in terms of maximum trapping force and maximum stably levitated object size, making the specific curvature and focal geometry a genuinely important design parameter rather than a cosmetic detail, and one that is carefully calculated (typically so that every individual transducer element's normal axis points at the same focal point) rather than chosen arbitrarily.

Element spacing and array density represent a related and equally important design consideration, particularly for phased-array systems intended for the more sophisticated holographic manipulation described earlier in this article. Published research on holographic acoustic tweezers has strongly suggested that the maximum achievable trapping forces are obtained when the array satisfies what is effectively a spatial sampling requirement analogous to the Nyquist criterion familiar from signal processing: the spacing between adjacent transducer elements needs to be small enough, relative to the acoustic wavelength, to avoid generating unwanted secondary interference patterns (sometimes called grating lobes, borrowing terminology directly from antenna and radar array design, where an analogous spatial sampling requirement applies) that would otherwise weaken and destabilize the intended primary trap. This is a genuinely quantitative design constraint, not a rule of thumb, and it directly determines how many individual transducer elements are needed to populate an array of a given physical size, which in turn determines the total electronics complexity, cost, and control-channel count required for a given system.

Field characterization and measurement, finally, is an aspect of acoustic levitation system design that is easy to overlook in a purely conceptual description but that matters enormously in practice because a levitator that has not been properly characterized is one whose actual performance is essentially unknown until an object is dropped in and either successfully traps or does not. Researchers characterize acoustic levitation fields using a range of measurement techniques, most commonly a calibrated microphone or hydrophone-style pressure sensor that can be precisely positioned within the working volume to map out the actual pressure field produced by a given array and driving configuration, allowing a direct, empirical comparison between the predicted field (calculated from the Gor'kov-based and multi-transducer models described earlier in this article) and the field the physical hardware actually produces, which will inevitably differ somewhat from the idealized calculation due to manufacturing tolerances in the transducers, minor misalignments in the 3D-printed mounting structure, and other real-world imperfections. More sophisticated characterization techniques, including laser Doppler vibrometry (which measures the minute vibrations of a target surface, including the surface of a levitated object itself, using the Doppler shift of reflected laser light) and high-speed optical imaging of levitated test particles as they are deliberately perturbed and observed returning to equilibrium, allow researchers to directly measure the stiffness of the trapping potential — effectively, how strongly and over what distance the restoring force acts — providing a direct, empirical validation of the theoretical trapping strength predicted by the underlying Gor'kov-based calculations, and giving practical engineers confidence that a given design will perform as intended before committing to a larger or more expensive build.

Taken together, these design details reinforce a point made throughout this article: acoustic levitation today is not a matter of a single clever trick or a single critical breakthrough, but rather the accumulated result of a well-developed engineering discipline, in which frequency selection, array curvature, element spacing, and field characterization are all understood quantitatively, published openly in the peer-reviewed literature, and available to be applied directly by anyone building a new system, rather than needing to be rediscovered through trial and error. This maturity, as much as any single hardware or software advance, is what makes it accurate to say that acoustic levitation is genuinely possible today, using knowledge, components, and methods that already exist and are already documented in detail.

Where the Work Is Happening

It is worth naming, concretely, where the advances described throughout this article have actually come from, both because it grounds the discussion in real, verifiable institutions rather than an abstract “researchers say,” and because it gives anyone wanting to go deeper a clear starting point for further reading.

The University of Bristol's Department of Mechanical Engineering, particularly the research group associated with Bruce Drinkwater and, in earlier stages of his career, Asier Marzo (now leading his own research group at the Public University of Navarre in Spain), has been the single most consistently productive source of the practical, accessible acoustic levitation advances described in this article. This group's contributions include the TinyLev open-source levitator design, the Ultraino open-source phased-array platform, the 2018 acoustic tractor beam work demonstrating stabilized large-object levitation using rapidly alternating acoustic vortices, and the 2019 holographic acoustic tweezers publication demonstrating independent multi-object manipulation using dual 256-element phased arrays. The group's consistent emphasis on open-source hardware and software design, publishing complete part lists, circuit schematics, 3D-printable design files, and control software alongside their peer-reviewed papers, is a large part of why acoustic levitation has become as broadly replicable as it has, and it reflects a deliberate philosophy on the group's part of prioritizing accessibility and reproducibility alongside pure scientific novelty.

The Institute of Science and Technology Austria, where Scott Waitukaitis (who began working on acoustic levitation as early as 2013) now leads a research group, has been responsible for some of the most recent advances in multi-object manipulation, including the late-2025 work on combining acoustic trapping with controlled electrostatic charging to prevent clumping between simultaneously levitated objects, discussed earlier in the section on cutting-edge developments. This group's work has tended to emphasize the more fundamental physics of multi-particle acoustic interactions — including the Bjerknes-force-driven clumping effects mentioned earlier — bringing a somewhat more foundational-physics orientation to the field alongside Bristol's more applications- and accessibility-oriented emphasis.

Beyond these two groups, acoustic levitation research is genuinely global and spans a wide range of institutional contexts. Materials-science-oriented containerless-processing work, building on the tradition established in the 1970s and 1980s, continues at various national laboratories and university materials-science departments with historical expertise in high-temperature melt processing. Biomedical and pharmaceutical applications of acoustic levitation, including the drug-formulation crystallization studies and targeted-drug-delivery research discussed earlier in this article, have been pursued at a range of pharmaceutical-science and biomedical-engineering departments internationally, often in institutions with a pre-existing strength in ultrasound-based medical technology, since much of the underlying phased-array and transducer expertise directly overlaps with the technology used in diagnostic and therapeutic medical ultrasound. Space-agency-affiliated research, following the historical thread established by Taylor Wang's Space Shuttle experiments, continues in a more limited but ongoing way at agencies and affiliated research institutions with an interest in containerless processing and biological research relevant to spaceflight.

Notably, and importantly for the overall argument of this article, a meaningful and growing fraction of current acoustic levitation activity now takes place entirely outside formal academic institutions altogether, in the hands of hobbyists, independent makers, and educators building TinyLev-derived and AcoustiFly-derived designs from openly published instructions, a genuinely new category of participant in the field that simply did not exist as a practical possibility before the low-cost, low-voltage, mass-producible component design philosophy pioneered by the Bristol group took hold in the mid-2010s. This diffusion of capability outside traditional research institutions is, in its own right, one of the more remarkable and under-appreciated developments in the recent history of the field, and it is a direct, practical demonstration of the central claim running through this entire article: acoustic levitation is not a technology confined to a small number of specialized laboratories, but one that has become genuinely, broadly achievable using knowledge and hardware already available today.

Can we do it?

Acoustic levitation is not a technology waiting to be invented. It is a technology that already exists, grounded in physics worked out across nearly a century of theoretical and experimental research, and built today from hardware — cheap piezoelectric transducers, low-voltage digital driver electronics, consumer 3D-printed structural components, and increasingly capable, freely published control software — that has become genuinely accessible to anyone with modest resources and a willingness to follow well-documented, peer-reviewed, and independently replicated designs. The question this article set out to answer, how acoustic levitation could be made to work today, has a concrete and satisfying answer at every level: the physics is understood well enough to be predictive rather than merely descriptive, the hardware needed to generate the required sound fields is commercially available and inexpensive, and the specific engineering designs needed to assemble that hardware into a working levitator have been published, tested, and replicated by research groups and hobbyists around the world.

What makes this particular moment in the technology's development genuinely interesting, rather than simply a continuation of decades-old laboratory practice, is the convergence of three trends: the democratization of the core hardware components through mass-produced, low-cost transducers originally developed for unrelated industries; the maturation of digital phased-array control techniques that transform acoustic levitation from a fixed, single-point phenomenon into a fully programmable, holographic manipulation tool capable of handling multiple objects simultaneously with precise positional and even rotational control; and a growing, increasingly sophisticated set of demonstrated applications across pharmaceuticals, materials science, biomedicine, electronics manufacturing, and space research that give the technology genuine practical purpose beyond its considerable value as a physics demonstration.

At the same time, an honest accounting of the field's real limitations is essential to keeping expectations properly calibrated. Acoustic levitation, as it exists today and as it is likely to exist for the foreseeable future, is fundamentally a technology for the precise, programmable, contact-free manipulation of small-to-modest objects, liquids, and biological samples — not a pathway toward levitating dense, human-scale objects, a goal that runs directly into hard physical limits on achievable acoustic force relative to gravitational load, compounded by genuine safety constraints on sound exposure that cannot simply be engineered away. Within its real and substantial niche, however, acoustic levitation offers a genuinely unique capability: contact-free, containerless, material-agnostic manipulation that no competing levitation technology — magnetic, electrostatic, aerodynamic, or optical — can match across the same combination of scale, material generality, and precision.

For a gravitational physicist, there is something particularly satisfying about a technology that makes the abstract balance of forces at the heart of gravitational physics so directly, visibly tangible: a small sphere hovering motionless in midair, held in place not by magic or by any exotic manipulation of gravity itself, but by the same everyday phenomenon of sound that carries a voice across a room, concentrated, shaped, and precisely balanced against the steady, unchanging pull of Earth's gravity. That balance, once purely a subject for advanced physics texts and specialized laboratories, is now something that can be built, observed, and understood firsthand by anyone with the interest and modest resources to do so — which is, in the end, the clearest possible demonstration that acoustic levitation is not a technology of tomorrow, but a technology of today.

For anyone reading this article who wants to move from understanding the physics to actually seeing it work, the practical path forward is clearer than it has ever been. It does not require a university affiliation, a specialized grant, or access to a national laboratory. It requires a modest budget for a few dozen ultrasonic transducers, access to a 3D printer (whether owned personally or borrowed at a local library or community makerspace), basic soldering skills that can be learned from any number of freely available tutorials, and the willingness to follow openly published, peer-reviewed, and independently replicated build instructions rather than starting from a blank page. The same is true, at a greater but still bounded level of effort, for anyone wanting to go further and build a genuinely programmable, holographic, multi-object phased-array system, following the open designs published by the research groups named in the previous section. What was once the exclusive province of specialized acoustics laboratories has become, through the steady, cumulative progress of transducer manufacturing, digital control electronics, and desktop 3D printing, a technology that sits within reach of a well-motivated individual working from a kitchen table or a home workshop. Sound, it turns out, was capable of holding matter aloft against gravity all along; what has changed, decisively and only in the last decade or so, is that the tools needed to harness that capability are finally, genuinely, and verifiably within everyone's reach.

That, in the end, is the real substance behind the question posed at the outset of this article. Acoustic levitation is possible today not because of some singular, dramatic discovery still on the horizon, but because a century of steady, cumulative, well-documented physics and engineering has quietly arrived at a point where the answer to “could this be built” has become, simply, “yes, and here is exactly how.”