Acoustic streaming is a representative nonlinear acoustic phenomenon, which is known as "sound-driven acceleration of medium". This effect is prominent only when the intensity of propagating sound is 120 dB SPL or greater. When an airborne ultrasound beam is used to generate this effect, straight, narrow ultrasound-driven airflow is produced, which can be highly localized within the high-intensity region of the ultrasound propagation.

Conventionally, fans or jets have been used to generate airflow. Acoustic streaming differs from those methods in that acoustic power at a certain position directly accelerates the air there, which results in advantageous physical characteristics. For example, the starting point of airflow by fans of jets is fixed to where those apparatuses are, meaning that the generated airflow decelerates and spreads as it travels further. In contrast, acoustic streaming can theoretically set its starting point fairly distant from the ultrasound beam emission source with a sufficiently large emission aperture, without being spread or blurred (Hasegawa+ 2017, 2019). In other words, a sufficiently focused airborne ultrasound beam can generate a laser-beam-like, straight, narrow airflow at an arbitrary position in a space.

Based on this physical property of acoustic streaming, we set the construction of ultrasound-driven airflow fields with functional spatiotemporal characteristics as a fundamental research issue. We also address the construction of related application systems, such as remote transportation and presentation of airborne fragrance (Hasegawa+ 2018) or an aerial barrier formed by acoustic streaming that suppresses direct exposure to exhaled infectious aerosols (Nagata+ 2024). We set our ultimate goal to the construction of a safe, comfortable, and functional indoor air environment that brings benefit to individual users anywhere in the room.

Sound waves are often used for non-contact measurement or testing, which is also the case with ultrasound. Major application scenarios include biomedical imaging and nondestructive testing, most of which just utilize linear sound properties such as reflection, diffraction, or scattering. We are particularly interested in utilizing nonlinear phenomena exhibited by intense ultrasound propagation. Such specific acoustic effects reflect particular physical properties of testing targets.

The Doppler effect is a major velocimetry methodology, where the observed frequency variation of waves emitted from the target is used to determine a particular velocity component of it. Although this method is widely used, the observed frequency change depends on the observation position. We newly found a similar effect (which may be referred to as the "parametric Doppler effect"), where the frequency variation of the parametric sound is related to a velocity component independent of where it is observed (Kotoku+ 2024).

We also devised a visualization method of audible sound fields, where only a single microphone is exploited to capture the parametric sound resulting from amplitude modulation of the target sound field and a spatially scanned ultrasound focus (Kozuka+ 2025). While most sound field visualization methods rely on the use of multi-channel recordings using many microphones, our method realizes the measurement directivity by a spatially localized ultrasound focus that turns the target audible sound into ultrasonic parametric sound, which allows for a visualization of a sound field by a monoaural recording. These measurement methodologies achieve the spatial selectivity not by increasing the number of sensor elements but by physically affecting the target field in a very spatially pinpoint manner.

The advent of the phased array system of air-coupled ultrasound transducers (devised by Prof. Shinoda from the univ. of Tokyo in 2008) has established a new research discipline of "midair nonlinear acoustics", where electronically controllable midair nonlinear acoustic effects are employed for various applications. This device was originally invented to realize a noncontact vibrotactile display. Nowadays, the application scope has been broadened to include new scenarios, such as real-time acoustic levitation of lightweight objects, steerable parametric sound presentation, etc. However, the phased array system needs a massive control circuit for driving individual transducers independently. This technical requirement inevitably makes the phased array system costly, and thus it is difficult to construct a large aperture of ultrasound emission planes. As a result, most current midair ultrasound application systems are tabletop ones with a small workspace. If one wants to create a whole-room-scale midair ultrasound application system, a much larger ultrasound emission aperture will be needed, which can form a sufficiently focused ultrasound field no matter how distant from the device. If such ultrasound sources are to be made in an inexpensive manner, nonlinear aerial ultrasound effects such as acoustic streaming, parametric arrays, or acoustic levitation will be available anywhere in the room.

We are working on the construction of such scalable convergent ultrasound emission sources that do not rely on the phased-array principle. Currently, we have established a method where an ultrasound emission body with fixed spatial emission patterns is partially covered with an "amplitude mask" that has a specific transmission distribution (Kitano+ 2023 , Hasegawa+ 2025 , Hasegawa 2026). This approach is originally known as the Fresnel Zone Plate (FZP), but we extended this framework to handle non-uniform incident waves to deform their wavefront to result in a desired convergent ultrasound field by a specifically designed amplitude mask. The resulting midair ultrasound patterns can be switched depending on the relative position of the mask and source, by which instantaneously reconfigurable ultrasound field formation was realized. This approach is quite simple, inexpensive, yet effective. We also aim to provide an alternative, affordable path to the construction of midair ultrasound application systems that is open to many researchers without access to the expensive and complicated phased array system.

Nonlinear acoustic effects include acoustic streaming (flow by sound), acoustic radiation force (mechanical force on solids or liquids by sound), or parametric sound (secondary sound emission by nonlinear acoustic interference). Our team makes full use of those effects to establish new application systems. We do not just think about how to apply a specific effect, but how the essential physics of those effects can be integrated into a new system.

The following are examples of recent application scenarios:

• Acoustic levitation of spherical objects larger than the wavelength using multiple vortex beam pairs

Midair ultrasonic manipulation based on acoustic radiation force has been a challenging research issue. We established a solution to this issue where two pairs of ultrasonic vortex beams form a target trapping region with a centripetal restoring force field that enables stable manipulation of a multi-wavelength spherical target (Momoki+2026). This achievement will lead to a new technique for handling lightweight multi-wavelength targets freely manipulated in the air.

• Personal 3D audio presentation based on direction-controlled parametric sound emission

The self-demodulation effect of amplitude- (or frequency-) modulated ultrasound has been known as parametric sound and used for super-directional audible sound presentation. We found that its subjectively perceived sound source direction and position (here we refer to this as sound image) can be drastically varied by altering the emission angle of ultrasound beams and experimentally constructed a psychophysical model to explain their relationship (Momoki+2025). The experiments show that fairly small individual differences are observed in the sound image perception, which exhibits the possibility of the construction of personalized 3D audio systems using sparse multiple parametric sound emitters.  

• Mode-selective sloshing excitation of liquids in a container by a spatiotemporal radiation force field on the liquid surface

Acoustic radiation force is used to remotely exert mechanical force. However, the resulting force is as small as several tens of millinewtons, which is valid only for actuation of extremely lightweight objects. At the same time, a good spatiotemporal localization unique to this nonlinear acoustic effect can be applied to large-scale actuation of a resonant target not necessarily lightweight. We demonstrated that water weighing several hundreds of grams in an open container can be rocked contactlessly with a large sloshing amplitude (Hijikata+2024).

• Midair push-button with two-stage tactile feedback by midair converged ultrasound giving enhanced believability of input completion

The ultrasound phased array system was originally devised to generate noncontact tactile sensation on the skin surface using radiation force. Nowadays, an emerging scientific discipline of Midair Haptics has been established among several research groups worldwide. A common tactile component that we come across is a push-button, where the believability of input completion by touch feeling plays an important role. We demonstrate that a two-stage ultrasonic pulse presentation on the user's fingers at the moments of touching the virtual button surface and lifting the finger off the button significantly enhances the believability of input completion (Sugawara+2025).