An array of thiophene-based monomers has been synthesized, with tailormade side groups and energetics characteristics, enabling cell- and tissue integration and in vivo-polymerization using enzymatic-, electro- and photo-polymerization protocols. Polymerized structures forming electrodes, wires, and simple device architectures have successfully been achieved onto and into individual cells, nerves, the central nervous system (CNS), and around the heart of different animal models to form electrode functionality for stimulation and recording exhibiting minimal levels of invasiveness. Our findings open for a new pathway of electroceuticals targeting disorders and degenerative diseases that combine pharmaceutical functionality with bioelectronics which is amalgamated with cells, organs, and CNS.

I introduce flow zoometry, a whole-animal equivalent of flow cytometry, designed for large-scale, individual-level, high-content screening of animals. This approach addresses the fundamental bottlenecks in animal research: high costs, labor-intensive processes, time inefficiency, susceptibility to human error, and limited statistical significance. Flow zoometry marks a paradigm shift in high-throughput screening, extending the progression from molecules to cells and now from cells to entire animals.

In the past decade the photonic community witnessed a complete transformation of optics. We are now able to define and control the flow of light using thousands of monolithically integrated optical components – all on a silicon chip. The main drive for silicon photonics is the ability to transmit and manipulate ultra high bandwidth with low power dissipation. Today there are hundreds of products being developed and commercialized towards this goal. The field of silicon photonics is rapidly evolving and is now enabling completely new applications. I will discuss these emerging applications, as well as the challenges of the field.

In scientific research, there is a growing need for manipulation and automation of micro- and nano-scale objects. Particularly in the biomedical field, genetic analysis technology is advancing rapidly, and the objects to be analyzed are now at the single-cell level. For analytical purposes, this scale is often used for scientific exploration, such as the investigation of unknown properties of living cells and tissues, and requires precise manipulation techniques that take into account the interaction with the fluid environment under analysis. We are working on robotics to expand capabilities in the microscopic world for biomedical innovations. Micro-nano mechatronics plays an important role in realizing new functions. Based on this approach, we investigated new capabilities of integrating robotic and microfluidic technologies and applied them to several scientific experimental tasks. For automation of scientific experiments purposes, micro-objects such as cells must be individually managed from the first three-dimensional space. Multi-scale operations on the order of the sixth power of 10 (micrometer to meter) should be realized. Automation through such micro/nanoscale operations is difficult and many challenges remain. We have realized the individual management and wide-are movement of each object by integrating a microfluidic chip into the end-effector of a robotic manipulator. We developed the associated technologies required for the automation of micro and nano works. Furthermore, microscopic manipulation is also important in tissue sampling within the body. For example, tissue sampling within the digestive system. This talk will introduce emerging robotic technologies expanding capabilities in the microscopic world, especially manipulation and automation at small scales in the biomedical field, and discuss future prospects.

Sensing the deformation of soft materials, soft robots, and textiles has many applications to human-machine interfaces, soft wearable devices for human health, and robotics. An existing challenge in designing these sensors is achieving high sensor performance through viscoelastic materials and translating them from the lab to daily use in a wide range of applications.

I will discuss sensor design challenges for these soft applications and present recent work that applies these concepts to origami robots, grippers, and wearable devices. I will also present work in enhancing the stability and mechanical selectivity of stretchable sensors and discuss applications for such sensors in wearable healthcare applications, soft robotics, and beyond.

Developing advanced transducer devices relies on a combination of standard cleanroom techniques—such as lithography, CVD, PVD, etching, bonding, and packaging—complemented by emerging, less conventional methods that expand the possibilities for materials, design, and integration in advanced micro- and nanosystems.

This talk will present recent research examples from our laboratory and collaborators, aimed at expanding the micro- and nanofabrication toolbox for sensor and actuator development. We will revisit emerging microscale techniques such as drop-on-demand inkjet printing (DOD IJP), two-photon 3D printing (2PP), and melt-electro writing (MEW), highlighting their relevance for MEMS/NEMS applications. Additionally, we will explore emerging nanoscale techniques, such as thermal scanning probe lithography (tSPL), with a focus on its recent advancements in grayscale patterning and deterministic strain engineering in 2D materials, aimed at tuning their semiconducting properties.

Beyond fabrication, education in micro- and nanomanufacturing is becoming increasingly critical. I will also present preliminary results from a pilot study exploring mixed-reality learning methodologies in the EPFL cleanroom, designed to enhance process engineering training. Looking ahead, fostering collaborations among universities and research centers will be essential in shaping the future of digital education in STEM fields.

In the quest for enhanced sensitivity and accuracy in sensor technology, traditional MEMS sensors face inherent limitations that necessitate the exploration of advanced alternatives. Quantum technologies have emerged as a revolutionary field, promising unprecedented advancements in various domains, particularly in sensing applications. By leveraging qubits to interrogate the environment, quantum sensors offer significant advantages over classical counterparts, including unprecedented precision and sensitivity.

One promising approach in quantum sensing involves the use of alkali vapor cells with xenon as gyroscopes. These gyroscopes operate based on the principles of atomic spin precession, where the interaction of alkali atoms with xenon gas allows for highly accurate measurements of rotational motion. The system architecture typically includes a laser to polarize the atoms, a magnetic field to induce precession, and detectors to measure the resulting signals. Applications of these gyroscopes span from navigation systems in autonomous vehicles to precision instrumentation in aerospace.

Another cutting-edge quantum sensing technology involves nitrogen-vacancy (NV) centers in diamond as magnetometers. NV centers are defects in the diamond lattice that exhibit remarkable sensitivity to magnetic fields while giving full access to the vectorial information of the magnetic field. The working principle involves optically detecting the spin state of the NV centers, which changes in response to external magnetic fields. The system architecture includes a laser for spin initialization and readout, microwave sources for spin manipulation, and photodetectors for signal acquisition. There are numerous applications of highly sensitive NV center magnetometers ranging from medical uses cases like magnetocardiography to EV battery monitoring and geological exploration.

The journey from laboratory research to industrialization of quantum sensors presents several challenges and success factors. At Bosch, we have established an internal startup to drive the development and commercialization of quantum sensors. Key challenges include ensuring the robustness and reliability of quantum devices, shrinking size and power demand, and scaling up production processes to achieve competitive cost structures. Our market strategy focuses on identifying high-impact applications devoted to our philosophy invented for life. In conclusion, quantum sensors represent a transformative leap in sensing technology, offering capabilities that were previously unattainable with classical approaches. The advancements in alkali vapor cell gyroscopes and NV center magnetometers demonstrate the practical potential of quantum sensing. As we continue to address industrialization challenges, the integration of quantum sensors into daily life is becoming increasingly feasible. The future outlook for quantum sensors is promising, with ongoing research and development paving the way for new applications and enhanced performance across various industries.

This report discusses the latest advancements in Artificial Intelligence (AI) enhanced sensors and their integration into Artificial Intelligence of Things (AIoT) sensing systems for cloud AI and edge AI enabled applications. Traditional sensor designs can no longer meet the demands for real-time sensing and large-scale data processing, leading to a shift towards a new paradigm of AI Sensors and AIoT sensing systems with integrated computational intelligence. The development of microelectromechanical systems (MEMS) and nanophotonic sensors provides a foundation for advancements in sensor technology, while breakthroughs in nanomaterials, heterogeneous integration, and novel structures will drive the next generation of AI-enhanced smart sensors.

Machine learning-enhanced nanophotonic sensors have improved gas detection and molecular recognition sensitivity and specificity. Hook-shaped nanostructures (HNAS) and wavelength-multiplexed structures (WMHNAS) combined with microfluidics enable high-sensitivity, label-free molecular recognition. AI-enhanced photonic nose platforms, incorporating plasmonic structures and metal-organic frameworks (MOFs), show potential in environmental monitoring and smart agriculture.

The development of MEMS vibration energy harvesters (VEHs) has enabled self-powered AIoT systems. Piezoelectric harvesters (P-VEHs) with features of high energy density and small form factors are expanding VEH frequency ranges and improving energy conversion efficiency. We have developed self-powered AIoT sensing nodes (ICUPE), successfully applied in harsh environment, transportation, and Industry 4.0, achieving high-precision acceleration, frequency, and tilt angle recognition.

In self-powered wearable AI sensors, advanced material innovations have achieved breakthroughs in flexibility, stretchability, transparency, and biocompatibility, promoting low-power or battery-free sensor platforms for applications like human-machine interaction, soft robotics, and electronic skin (e-Skin). Triboelectric nanogenerator (TENG) based on textiles have shown great potential in sensing applications such as gesture recognition, virtual reality (VR) and augmented reality (AR). Combined with convolutional neural networks (CNNs), these sensors enable high-precision gesture recognition.

With the growing demand for real-time and energy-efficient AI applications, the rapid development of edge computing and edge AI at the sensor nodes has led to the emergence of "in-sensor computing." This approach embeds computational capabilities within sensors, reducing data transmission burdens while enhancing privacy and energy efficiency. As optical edge computing platform advanced by Si Photonics, optical AIoT systems have become an attractive approach for enabling intelligent, energy-efficient, adaptive sensing networks.

By combining cloud and/or edge computing technologies, diversified smart sensors provide new solutions aiming at personalized healthcare, smart homes, smart cities and VR/AR applications. These innovations will revolutionize our living environment and enable efficient, energy-saving, privacy-protected and responsive intelligent environments.

Microfabrication technologies have been utilized for the construction of functional microstructures with large volume and low cost, which has led to investigations on design, material, packaging, and integration for numerous applications in sensors, actuators, and microsystems. In this presentation, I will discuss our works based on both piezoelectric and piezoelectret materials. First, leveraging from our experiences in the field of MEMS (Microelectromechanical Systems), we have been working on piezoelectric micromachined ultrasonic transducers (PMUTs) based on AlN, Lithium Niobate, and KNN ((K,Na)NbO) thin films toward practical applications, including: (1) flow sensors, (2) catheter sensors, (3) blood flows sensors, (4) structural curvature sensors, (5) surface temperature sensors; (6) 3D space imaging, (7) directional loudspeakers, (8) haptic sensations, (9) fluid property sensors, (10) temperature, humidity and pressure sensors; (11) bone age sensors ... etc. The fundamentals and prospects will also be discussed. The second part of this presentation will focus on the development of piezoelectret materials for sensors and actuators. In a prototype wearable device based on fluorinated ethylene propylene (FEP) thin films, its actuator mode can generate force as high as 20 mN - comparable to a common cell phone in the vibration mode. The sensor mode can detect as light as a dandelion seed and this prototype system has an ultrahigh piezoelectric coefficient with low driving voltage to outperform common devices made of piezoelectric materials. The sensing function has been utilized for real-time and continuous monitoring of physiological signals such as human pulses to conduct the pulse palpation measurement like well-trained doctors in Traditional Chinese Medicine and the heart, breath, and Korotkoff sounds for various health monitoring applications. The actuator function has been utilized for haptic feedback applications. Finally, I will introduce some fun robotics projects, such as an ultra-robust and fast moving piezoelectric robot similar to those of cockroaches and the smallest untethered flying robot driving in a magnetic field.

The recent developments in nanotechnology have paved the way for the generation of a variety of nanostructures, crucial for high-performance biosensors, terahertz (THz) devices, and metadevices. By integrating nanoimprint technology with precise dry etching techniques, it is possible to produce three-dimensional (3D) nanodevices while maintaining high uniformity and accurate dimensional control over large areas. This presentation will highlight several applications of these advanced technologies: 3D biomimetic platforms and plasmonic biosensors designed for the precise control and highly sensitive monitoring of cells and biomolecules; high-frequency THz lenses and antennas that leverage curved or meta-surfaces to boost functionality and performance; and multiple-layered meta surfaces with twist angles engineered to achieve chiral magic angles for sophisticated light manipulation.

With the advent of the 4th industrial revolution, there has been a significant rise in the use of Internet of Things (IoT) and advanced sensors in several sectors, including but not limited to smart factories, healthcare, environment, and entertainment. However, the deployment of an increasing number of sensors in IoT systems has made it crucially important but challenging to reduce their form factors and power consumption. Consequently, there is a growing need for sensors that are miniaturized, low-power, or self-powered. In particular, when it comes to environmental IoT, it is essential to decrease power consumption in gas sensors while simultaneously enhancing sensing performance factors, such as sensitivity, selectivity, and response speed. In this talk, we present some recent advancements in the development of environmental sensors that are miniaturized, low-power or self-powered. These sensors utilize functional nanomaterials, microfabricated sensor structures, and photo-activated sensing mechanisms, which were developed by our research group at KAIST. Additionally, we describe the implementation of deep learning-based signal processing techniques for ultra-low power environmental sensors. The following topics are covered in depth: (a) Low-power semiconductor metal oxide (SMO) based chemoresistive gas sensor array using localized hydrothermal synthesis, (b) Ultra-low-power SMO gas sensors monolithically integrated on a micro LED platform, (c) Self-powered gas sensors using chemo-optically modulating sensing films and photovoltaic cells, (d) Machine learning of gas sensor array for high performance electronic nose (e-nose) systems.

Micron-scale swimming robots or “microswimmers” have enormous potential for use in biomedical applications such as drug delivery, diagnostics and even microsurgery. However, to realize their potential to locomote, do work, and adapt to delicate environments, these microscale actuators must meet strict engineering requirements such as monodispersity, biocompatibility, multi-functionality and modularity. In this talk, I will highlight our recent progress towards the experimental realization of magnetic colloidal microswimmers with flexible linkages. The energy for actuation is provided by a constant magnitude oscillating magnetic field that imparts an oscillating torque on the magnetic components of the microswimmers. In this low Reynolds environment, reciprocal motion of a rigid or rigidly-fused magnetic system does not promote locomotion. However, when flexible linkages are introduced between pairs of particles, that compliance results in non-reciprocal traveling wave-type motion between linked particles, which overcomes the Scallop Theorem and enables productive locomotion [1].

In our first demonstration of this top-down and bottom-up approach, we utilized two-photon polymerization and polydimethylsiloxane (PDMS) molding to microfabricate templates with arrays of hemispherical pockets [2]. Self-assembly of the microspheres in the templates was performed using templated assembly via selective release (TASR), which achieved controlled particle placement and spacing, with enhanced size-selection of particles [2]. Biocompatible and flexible DNA nanotubes were used to connect the streptavidin-coated colloidal particles while accommodating a range of micron-scale particle separations. These DNA nanotechnology linkers provide compliance and the potential for both reconfigurability and chemical responsiveness. To enable modular and multifunctional microswimmers, we utilized high surface energy polycarbonate templates fabricated using polycarbonate heat (PCH) molding [3]. This change in template material resulted in a 100x increase in pocketing efficiency, enabling the fabrication of larger populations with increasingly complex body plans. The triggered separation of microswimmer particles also demonstrates the potential for chemically-responsive behaviors such as on-demand cargo release.

To understand the impact of body plan and particle size and shape on locomotion, we used two-photon polymerization to create flexible millimeter-scale models of the magnetic microswimmers for systematic investigation of actuation-structure-locomotion relationships [4]. These models provide a high level of structural monodispersity, with locomotion studies revealing actuation frequencies for maximum speed and dependence of swimming speed on linker stiffness and particle aspect ratio [4]. These studies will inform improved designs for microswimmers and support the generation of improved computational models of microswimmer locomotion. Lastly, to enable assessment of microswimmer locomotion in complex flow environments, we have also developed a methodology that extracts motion due to actuation from motion due to local fluidic disturbances [5].

I’ll end the talk with a discussion of materials for microswimmers, highlight emerging opportunities for using nanostructured microparticles, synthetic cells, and xeno nucleic acids to enable transformative applications in transport and sensing in the lab, in vitro and beyond [6].

Global climate change threatens ecosystem functioning worldwide. Forest ecosystems are particularly important for carbon sequestration. However, recurrent stresses, such as heat waves, floods, and droughts, increasingly endanger forests worldwide, with potentially cascading effects on their carbon sink capacity and sustainability. Knowledge on the impact on the multitude of processes driving soil-plant-atmosphere interactions within these complex systems is widely lacking and uncertainty about future changes extremely high. Thus, forecasting forest response to climate change will require an improved understanding of carbon and water cycles across various temporal and spatial scales, from minutes to seasons, from leaves to ecosystem, covering the atmosphere, biosphere, pedosphere and hydrosphere. Many relevant processes occur at small scales and high spatial heterogeneity and their interactions can be key players to amplify or dampen a system’s response to stress. Currently, we are lacking the appropriate measuring, data and modelling tools allowing for comprehensive, real time quantification of relevant processes at high spatio-temporal coverage.

Consequently, the interdisciplinary research project ECOSENSE develops, implements, and evaluates a versatile, distributed, autonomous, intelligent sensor network based on novel microsensors tailored to the specific needs in harsh forest environments. All measurements need to happen in a minimally invasive manner in order not to disturb the ecosystem we are exploring. This requires consideration of size, weight, inertness, cleanliness, and robustness. The complex environment of a remote naturally structured forests with rain, wind, insects and mammals thereby poses its own challenges that the ECOSENSE sensors are facing.

The network comprises distributed miniaturized sensors for 1) the quantification of gas and water fluxes, specifically isotope discriminated CO2 and volatile organic compounds (VOC, here isoprene) at ppb concentrations, and 2) for the detection of typical stress markers, such as active, laser induced chlorophyll fluorescence. ECOSENSE thus explores these relevant parameters across all technical scales, i.e. from small distributed sensors on leaf level including low power electronics, to medium sized laser spectrometers, up to drones and eddy covariance towers. In our recently established field site, the ECOSENSE Forest, we are currently operating 3000 continuous data streams, partially with a temporal resolution of a few seconds. Measured data is transferred in real-time to a sophisticated data base and can be explored for process analysis and AI enhanced simulation models providing the basis for future predictions of process-based alterations in ecosystem functioning and sustainability.