Scientific Publications

Disturbed flow is a hallmark of diseased vasculature, yet its influence on particle behavior under external actuation remains poorly understood. We uncover distinct behaviors of microparticles under disturbed flow when exposed to ultrasound, revealing selective trapping and aggregation phenomena that differ fundamentally between soft and rigid particles. Using microfluidic models of disturbed vascular flow, we show that microbubbles become trapped at the eye of vortices and self-assemble via ultrasound-induced forces. As clusters grow to a critical size, they are ejected and adhere to the wall opposite the ultrasound source, forming nuclei that progressively occupy aneurysm cavities—a mechanism that could enable targeted, noninvasive treatment. These findings reveal an unexplored interplay between ultrasound and hydrodynamic forces, offering a new strategy for ultrasound-guided therapeutic delivery in vascular disease.

Imitating the shape-encoded tactics of natural microswimmers—organisms that flip, roll, and rheotaxis through viscous fluids—could transform microfluidics, micromanufacturing, and targeted therapy. However, translating those geometric navigation cues into actively driven microrobots is an open, largely unexplored challenge. Here, inspired by the structure of sperm cells, we introduce a sound-propelled head-helix microparticle (“microrobot”) featuring an elliptical head and a spiral tail. This asymmetrical design interacts with the incident acoustic field, generating complex secondary flows that induce a torque, enabling the particle to reorient around its cross-section. The microparticle exhibits a preferred direction of propulsion and orientation when exposed to a traveling sound wave, reorienting if its initial alignment deviates from this preference. Both the preferred direction and orientation can be modulated by adjusting the sound frequency, and they further adapt to background flow fields in the environment. Furthermore, the microparticle exhibits rheotaxis-like motion, exhibiting wall-following motion with frequency-dependent sliding behavior. By moving towards the channel wall, it enters the region with the smallest flow velocities, allowing it to move antiparallel to the fluid. These findings contribute to the engineering of the trajectories of sound-propelled microparticles and to the development of next-generation microrobots for medical and other innovative applications.

Acoustically actuated soft matter offers potential for agile microscale manipulation, yet acoustic-soft matter interaction at the microscale remains poorly understood. Here, we explore the mechanism of ultrasound-soft matter interaction by developing a bio-inspired ultrasound-driven soft hydrogel microgripper. This exploration allows to delve deeper into the understanding of nonlinear dynamics, mode coupling, and energy transfer. The developed microgripper (≤ 120 µm) overcomes key challenges of existing grippers, including complex fabrication, reliance on additives or external wiring, rigid structures, slow or poorly controllable responses, and risks of sample damage or contamination. Interacting with acoustic actuation, soft microgrippers oscillate and deform, while adjusting acoustic parameters and microgrippers’ structures allows for programmable interactions. The optimized acoustic actuation of the soft microgripper enables precise, ultrafast (∼2 ms) gripping and handling of distinct delicate objects. This work advances the integration of soft matter with acoustic actuation especially at the microscale, offering a versatile, reliable, and scalable solution for microrobotics, targeted drug delivery, and lab-on-a-chip applications.

Visualization and tracking of microrobots in real time pose key challenges for surgical microrobotic systems, as existing imaging modalities like magnetic resonance imaging, computed tomography, and x-ray are unable to monitor microscale items with real-time resolution. Ultrasound imaging-guided drug administration represents a remarkable advancement in this respect, offering real-time visual feedback on invasive medical procedures. However, ultrasound imaging still faces substantial inherent limitations in spatial resolution and signal attenuation, which hinder extending this method to microrobot visualization. Here, we introduce an approach for visualizing individual microrobots in real time with color flow mapping ultrasound imaging based on acoustically induced structural oscillations of the microrobot generating a pseudo-Doppler signal. This approach enables the simultaneous localization and activation of bubble-based microrobots using two ultrasound sources operating at distinct frequency bandwidths. Our successful capture of microrobots measuring 60 to 80 micrometers in diameter reveals the potential of real-time ultrasonic imaging at the microscale.

Inspired by Richard Feynman’s 1959 lecture and the 1966 film Fantastic Voyage, the field of micro/nanorobots has evolved from science fiction to reality, with significant advancements in biomedical and environmental applications. Despite the rapid progress, the deployment of functional micro/nanorobots remains limited. This review of the technology roadmap identifies key challenges hindering their widespread use, focusing on propulsion mechanisms, fundamental theoretical aspects, collective behavior, material design, and embodied intelligence. We explore the current state of micro/nanorobot technology, with an emphasis on applications in biomedicine, environmental remediation, analytical sensing, and other industrial technological aspects. Additionally, we analyze issues related to scaling up production, commercialization, and regulatory frameworks that are crucial for transitioning from research to practical applications. We also emphasize the need for interdisciplinary collaboration to address both technical and nontechnical challenges, such as sustainability, ethics, and business considerations. Finally, we propose a roadmap for future research to accelerate the development of micro/nanorobots, positioning them as essential tools for addressing grand challenges and enhancing the quality of life.

Reinforcement learning is emerging as a powerful tool for microrobots control, as it enables autonomous navigation in environments where classical control approaches fall short. However, applying reinforcement learning to microrobotics is difficult due to the need for large training datasets, the slow convergence in physical systems and poor generalizability across environments. These challenges are amplified in ultrasound-actuated microrobots, which require rapid, precise adjustments in high-dimensional action space, which are often too complex for human operators. Addressing these challenges requires sample-efficient algorithms that adapt from limited data while managing complex physical interactions. To meet these challenges, we implemented model-based reinforcement learning for autonomous control of an ultrasound-driven microrobot, which learns from recurrent imagined environments. Our non-invasive, AI-controlled microrobot offers precise propulsion and efficiently learns from images in data-scarce environments. On transitioning from a pretrained simulation environment, we achieved sample-efficient collision avoidance and channel navigation, reaching a 90% success rate in target navigation across various channels within an hour of fine-tuning. Moreover, our model initially generalized successfully in 50% of tasks in new environments, improving to over 90% with 30 min of further training. We further demonstrated real-time manipulation of microrobots in complex vasculatures under both static and flow conditions, thus underscoring the potential of AI to revolutionize microrobotics in biomedical applications.

Wearable electronics, such as smart textiles, are of potential use in healthcare monitoring, human-machine interfaces and environmental analysis. However, the scalability and reliability of the technology is restricted due to challenges related to rapid material degradation, potential toxicity, high production costs and heavy computational workload. Here we report an acoustic-based smart textile technology. The approach, which we term SonoTextiles, uses piezoelectric transducers that are mounted at both ends of glass microfibres and act as transmitters and receivers of acoustic waves. The flexible glass microfibres act as acoustic waveguides and are embedded into the textile substrate, providing precise sensing by measuring wave propagation and energy loss along the fibre in response to stimuli such as touch and bending. We also use acoustic frequency selectivity and frequency-domain signal processing algorithms to enhance computational efficiency. Our acoustic textile is breathable, durable and stable under thermal fluctuations, and we show that it can be used in distributed tactile sensing, hand gesture recognition and respiratory rate monitoring.

Biofilm formation and encrustation are major issues in indwelling medical devices, such as urinary stents and catheters, as they lead to blockages and infections. Currently, to limit these effects, frequent replacements of these devices are necessary, resulting in a significant reduction in patients’ quality of life and an increase in healthcare costs. To address these challenges, by leveraging recent advancements in robotics and microfluidic technologies, we envision a self-cleaning system for indwelling medical devices equipped with bioinspired ultrasound-activated cilia. These cilia could be regularly activated transcutaneously by ultrasound, generating steady streaming, which can be used to remove encrusted deposits. In this study, we tested the hypothesis that the generated streaming can efficiently remove encrustations and biofilm from surfaces. To this end, we developed a microfluidic model featuring ultrasound-activated cilia on its wall. We showed that upon ultrasound activation, the cilia generated intense, steady streaming, reaching fluid velocity up to 10 mm/s. In all our experiments, this mechanism was able to efficiently clean typical encrustation (calcium carbonate and oxalate) and biofilm found in urological devices. The generated shear forces released, broke apart, and flushed away encrusted deposits. These findings suggest a broad potential for ultrasound-activated cilia in the maintenance of various medical devices. Compared to existing methods, our approach could reduce the need for invasive procedures, potentially lowering infection risks and enhancing patient comfort.

Gas-filled microbubbles have been extensively used as contrast agents for ultrasound therapies and have recently been explored as microrobots. When exposed to an intense acoustic wave, microbubbles oscillate and scatter the sound field, leading to assembly and manipulation behaviors. Although traditional microbubbles demonstrated promising potential for manipulation in physiological environments, we recognized the need for improved assembly selectivity and acoustically directed propulsion. For this, we developed a new microbubble design using microfluidics. We engineered microbubbles with an outer oil droplet encapsulating a gas core, allowing the gas to move freely within the oil. When the gas reaches the droplet’s periphery, the oil attenuates the scattered wave, ensuring that scattering or acoustic amplification is concentrated near the droplet’s edge. This approach enabled us to control the organized assembly of microbubbles and direct their acoustic navigation. We believe this approach will open new possibilities for using gas-filled microbubbles as safer drug carriers or for micro-interventions in biomedical settings.

Navigating medical microrobots through intricate vascular pathways is challenging. AI-driven microrobots that leverage reinforcement learning and generative algorithms could navigate the body’s complex vascular network to deliver precise dosages of medication directly to targeted lesions.

Regenerative medicine continually seeks effective methods to address spinal cord injuries (SCI), which are known for their limited regenerative potential. Despite advances in neural progenitor cell (NPC) transplants for spinal cord injuries, challenges related to graft survival, reliable in vivo differentiation, and neural integration significantly hinder real functional recovery and limit clinical outcomes. This study introduces ‘NPCbots’, biohybrid microrobots engineered by integrating human-induced pluripotent stem cell-derived NPCs with magnetoelectric nanoparticles composed of cobalt ferrite-barium titanate. These enable magnetic navigation and neuronal stimulation, enhancing targeted therapeutic interventions. Our lab-on-a-chip system allows for the mass production of NPCbots, ensuring their differentiation and biocompatibility. Remarkably, in a zebrafish model of SCI, NPCbots stimulated by an alternating magnetic field demonstrated rapid in vivo differentiation and integration into damaged neural pathways, significantly enhancing neural regeneration. Within three days, injured zebrafish treated with NPCbots exhibited almost normal swimming behavior and significantly improved exploratory behavior, showcasing the potential of NPCbots to swiftly repair neural structures and restore the central nervous system’s functionality in spinal cord injury models through non-invasive means. Additionally, precise in vitro and in vivo manipulation of NPCbots indicates their broader application in various neurodegenerative disorders, offering a promising route for effective spinal cord and neurological recovery.

Advances in additive manufacturing in composites have transformed aerospace, medical devices, tissue engineering, and electronics. A key aspect of enhancing properties of 3D-printed objects involves fine-tuning the material by embedding and orienting reinforcement within the structure. Existing methods for orienting these reinforcements are limited by pattern types, alignment, and particle characteristics. Acoustics offers a versatile method to control the particles independent of their size, geometry, and charge, enabling intricate pattern formations. However, integrating acoustics into 3D printing has been challenging due to the scattering of the acoustic field between polymerized layers and unpolymerized resin, resulting in unwanted patterns. To address this challenge, SonoPrint, an innovative acoustically assisted volumetric 3D printer is developed that enables simultaneous reinforcement patterning and printing of the entire structure. SonoPrint generates mechanically tunable composite geometries by embedding reinforcement particles, such as microscopic glass, metal, and polystyrene, within the fabricated structure. This printer employs a standing wave field to create targeted particle motifs-including parallel lines, radial lines, circles, rhombuses, hexagons, and polygons-directly in the photosensitive resin, completing the print in just a few minutes. SonoPrint enhances structural properties and promises to advance volumetric printing, unlocking applications in tissue engineering, biohybrid robots, and composite fabrication.