Self-Powered Artificial Mechanoreceptor Based on Triboelectrification for a Neuromorphic Tactile System

Near-Sensor Edge Computing System Enabled by a CMOS Compatible Photonic Integrated Circuit Platform Using Bilayer AlN/Si Waveguides

The rise of large-scale artificial intelligence (AI) models, such as ChatGPT, DeepSeek, and autonomous vehicle systems, has significantly advanced the boundaries of AI, enabling highly complex tasks in natural language processing, image recognition, and real-time decision-making. However, these models demand immense computational power and are often centralized, relying on cloud-based architectures with inherent limitations in latency, privacy, and energy efficiency. To address these challenges and bring AI closer to real-world applications, such as wearable health monitoring, robotics, and immersive virtual environments, innovative hardware solutions are urgently needed. This work introduces a near-sensor edge computing (NSEC) system, built on a bilayer AlN/Si waveguide platform, to provide real-time, energy-efficient AI capabilities at the edge. Leveraging the electro-optic properties of AlN microring resonators for photonic feature extraction, coupled with Si-based thermo-optic Mach-Zehnder interferometers for neural network computations, the system represents a transformative approach to AI hardware design. Demonstrated through multimodal gesture and gait analysis, the NSEC system achieves high classification accuracies of 96.77% for gestures and 98.31% for gaits, ultra-low latency (< 10 ns), and minimal energy consumption (< 0.34 pJ). This groundbreaking system bridges the gap between AI models and real-world applications, enabling efficient, privacy-preserving AI solutions for healthcare, robotics, and next-generation human-machine interfaces, marking a pivotal advancement in edge computing and AI deployment.

Neuromorphic devices for electronic skin applications

Neuromorphic devices represent an important advancement in technology, drawing inspiration from the intricate and efficient mechanisms of the human brain. This review paper elucidates the diverse landscape of neuromorphic electronic skin (e-skin) technologies while highlighting their numerous applications. Here, neuromorphic devices for e-skin are classified as two types of direct neuromorphic e-skins combining both neuromorphic devices and sensors, and indirect e-skins separating neuromorphic devices and sensors. In direct neuromorphic e-skins, there are developing devices like memristor-based neuromorphic devices with sensors and transistor-based neuromorphic devices with sensors. On the other hand, indirect types are demonstrated as separated neuromorphic and sensor parts systems through the various interfacing structures. It also describes recent neuromorphic developments in artificial neural networks (ANNs), deep neural networks (DNNs), and convolutional neural networks (CNNs), for the real-time interpretation of sensory data. Moreover, it introduces multimodal sensory feedback, soft and flexible e-skins, and more intuitive human-machine interfaces. This review examines various applications, including smart textiles for the development of next-generation wearable bioelectronics, brain-sensing interfaces that enhance tactile perception, and the integration of human-machine interfaces aimed at replicating the biological sensorimotor loop, which can improve health monitoring and biomedical applications. Additionally, the review also highlights the potential of neuromorphic e-skin in human-robot interaction, particularly in the context of continuous prosthetic control and robotics. Through this analysis, the paper provides insights into current advancements, identifies key challenges, and suggests future research directions for optimizing neuromorphic e-skin devices and expanding their practical implementation.

Self-Powered Artificial Neuron Devices: Towards the All-In-One Perception and Computation System

The increasing demand for energy supply in sensing units and the computational efficiency of computation units has prompted researchers to explore novel, integrated technology that offers high efficiency and low energy consumption. Self-powered sensing technology enables environmental perception without external energy sources, while neuromorphic computation provides energy-efficient and high-performance computing capabilities. The integration of self-powered sensing technology and neuromorphic computation presents a promising solution for an all-in-one system. This review examines recent developments and advancements in self-powered artificial neuron devices based on triboelectric, piezoelectric, and photoelectric effects, focusing on their structures, mechanisms, and functions. Furthermore, it compares the electrical characteristics of various types of self-powered artificial neuron devices and discusses effective methods for enhancing their performance. Additionally, this review provides a comprehensive summary of self-powered perception systems, encompassing tactile, visual, and auditory perception systems. Moreover, it elucidates recently integrated systems that combine perception, computing, and actuation units into all-in-one configurations, aspiring to realize closed-loop control. The seamless integration of self-powered sensing and neuromorphic computation holds significant potential for shaping a more intelligent future for humanity.

Bionic Recognition Technologies Inspired by Biological Mechanosensory Systems

Mechanical information is a medium for perceptual interaction and health monitoring of organisms or intelligent mechanical equipment, including force, vibration, sound, and flow. Researchers are increasingly deploying mechanical information recognition technologies (MIRT) that integrate information acquisition, pre-processing, and processing functions and are expected to enable advanced applications. However, this also poses significant challenges to information acquisition performance and information processing efficiency. The novel and exciting mechanosensory systems of organisms in nature have inspired us to develop superior mechanical information bionic recognition technologies (MIBRT) based on novel bionic materials, structures, and devices to address these challenges. Herein, first bionic strategies for information pre-processing are presented and their importance for high-performance information acquisition is highlighted. Subsequently, design strategies and considerations for high-performance sensors inspired by mechanoreceptors of organisms are described. Then, the design concepts of the neuromorphic devices are summarized in order to replicate the information processing functions of a biological nervous system. Additionally, the ability of MIBRT is investigated to recognize basic mechanical information. Furthermore, further potential applications of MIBRT in intelligent robots, healthcare, and virtual reality are explored with a view to solve a range of complex tasks. Finally, potential future challenges and opportunities for MIBRT are identified from multiple perspectives.

Spiking Neural Network Pressure Sensor

Von Neumann architecture requires information to be encoded as numerical values. For that reason, artificial neural networks running on computers require the data coming from sensors to be discretized. Other network architectures that more closely mimic biological neural networks (e.g., spiking neural networks) can be simulated on von Neumann architecture, but more important, they can also be executed on dedicated electrical circuits having orders of magnitude less power consumption. Unfortunately, input signal conditioning and encoding are usually not supported by such circuits, so a separate module consisting of an analog-to-digital converter, encoder, and transmitter is required. The aim of this article is to propose a sensor architecture, the output signal of which can be directly connected to the input of a spiking neural network. We demonstrate that the output signal is a valid spike source for the Izhikevich model neurons, ensuring the proper operation of a number of neurocomputational features. The advantages are clear: much lower power consumption, smaller area, and a less complex electronic circuit. The main disadvantage is that sensor characteristics somehow limit the parameters of applicable spiking neurons. The proposed architecture is illustrated by a case study involving a capacitive pressure sensor circuit, which is compatible with most of the neurocomputational properties of the Izhikevich neuron model. The sensor itself is characterized by very low power consumption: it draws only 3.49 μA at 3.3 V.

Spike-Based Neuromorphic Hardware for Dynamic Tactile Perception with a Self-Powered Mechanoreceptor Array

A self-powered mechanoreceptor array is demonstrated using four mechanoreceptor cells for recognition of dynamic touch gestures. Each cell consists of a triboelectric nanogenerator (TENG) for touch sensing and a bi-stable resistor (biristor) for spike encoding. It produces informative spike signals by sensing a force of an external touch and encoding the force into the number of spikes. An array of the mechanoreceptor cells is utilized to monitor various touch gestures and it successfully generated spike signals corresponding to all the gestures. To validate the practicality of the mechanoreceptor array, a spiking neural network (SNN), highly attractive for power consumption compared to the conventional von Neumann architecture, is used for the identification of touch gestures. The measured spiking signals are reflected as inputs for the SNN simulations. Consequently, touch gestures are classified with a high accuracy rate of 92.5%. The proposed mechanoreceptor array emerges as a promising candidate for a building block of tactile in-sensor computing in the era of the Internet of Things (IoT), due to the low cost and high manufacturability of the TENG. This eliminates the need for a power supply, coupled with the intrinsic high throughput of the Si-based biristor employing complementary metal-oxide-semiconductor (CMOS) technology.

Age of Flexible Electronics: Emerging Trends in Soft Multifunctional Sensors

With the commercialization of first-generation flexible mobiles and displays in the late 2010s, humanity has stepped into the age of flexible electronics. Inevitably, soft multifunctional sensors, as essential components of next-generation flexible electronics, have attracted tremendous research interest like never before. This review is dedicated to offering an overview of the latest emerging trends in soft multifunctional sensors and their accordant future research and development (R&D) directions for the coming decade. First, key characteristics and the predominant target stimuli for soft multifunctional sensors are highlighted. Second, important selection criteria for soft multifunctional sensors are introduced. Next, emerging materials/structures and trends for soft multifunctional sensors are identified. Specifically, the future R&D directions of these sensors are envisaged based on their emerging trends, namely i) decoupling of multiple stimuli, ii) data processing, iii) skin conformability, and iv) energy sources. Finally, the challenges and potential opportunities for these sensors in future are discussed, offering new insights into prospects in the fast-emerging technology.

Recent Progress in Wearable Near-Sensor and In-Sensor Intelligent Perception Systems

As the Internet of Things (IoT) becomes more widespread, wearable smart systems will begin to be used in a variety of applications in people's daily lives, not only requiring the devices to have excellent flexibility and biocompatibility, but also taking into account redundant data and communication delays due to the use of a large number of sensors. Fortunately, the emerging paradigms of near-sensor and in-sensor computing, together with the proposal of flexible neuromorphic devices, provides a viable solution for the application of intelligent low-power wearable devices. Therefore, wearable smart systems based on new computing paradigms are of great research value. This review discusses the research status of a flexible five-sense sensing system based on near-sensor and in-sensor architectures, considering material design, structural design and circuit design. Furthermore, we summarize challenging problems that need to be solved and provide an outlook on the potential applications of intelligent wearable devices.

Biomimetic Exogenous "Tissue Batteries" as Artificial Power Sources for Implantable Bioelectronic Devices Manufacturing

Implantable bioelectronic devices (IBDs) have gained attention for their capacity to conformably detect physiological and pathological signals and further provide internal therapy. However, traditional power sources integrated into these IBDs possess intricate limitations such as bulkiness, rigidity, and biotoxicity. Recently, artificial "tissue batteries" (ATBs) have diffusely developed as artificial power sources for IBDs manufacturing, enabling comprehensive biological-activity monitoring, diagnosis, and therapy. ATBs are on-demand and designed to accommodate the soft and confining curved placement space of organisms, minimizing interface discrepancies, and providing ample power for clinical applications. This review presents the near-term advancements in ATBs, with a focus on their miniaturization, flexibility, biodegradability, and power density. Furthermore, it delves into material-screening, structural-design, and energy density across three distinct categories of TBs, distinguished by power supply strategies. These types encompass innovative energy storage devices (chemical batteries and supercapacitors), power conversion devices that harness power from human-body (biofuel cells, thermoelectric nanogenerators, bio-potential devices, piezoelectric harvesters, and triboelectric devices), and energy transfer devices that receive and utilize external energy (radiofrequency-ultrasound energy harvesters, ultrasound-induced energy harvesters, and photovoltaic devices). Ultimately, future challenges and prospects emphasize ATBs with the indispensability of bio-safety, flexibility, and high-volume energy density as crucial components in long-term implantable bioelectronic devices.

Skin-Friendly and Wearable Iontronic Touch Panel for Virtual-Real Handwriting Interaction

Touch panels are deemed as a critical platform for the future of human-computer interaction and metaverse. Recently, stretchable iontronic touch panels have attracted attention due to their superior adhesivity to the human body. However, such adhesion can not be named "real wearable", leading to discomfort for the wearer, such as rashes or itching with long-time wearing. Herein, a skin-friendly and wearable iontronic textile-based touch panel with highly touch-sensing resolution and deformation insensitivity is designed based on an in-suit growing strategy. This textile-based touch panel endows excellent interfacial hydrophilic and biocompatibility with human skin by overcoming the bottlenecks of the hydrogel-based uncomfortable sticky touch interface and low mechanical behavior. The developed touch panel enables handwriting interaction with good mechanical capacity (114 MPa), nearly 4145 times higher than pure hydrogel. More importantly, our touch panel possesses intrinsic insensitivity to wide external loading from the silver fiber (<0.003 g) to even heavy metal block (>10 kg). As proof of concept, the textile-based iontronic touch panel is applied to handwriting interaction, such as a flexible keyboard and wearable sketchpad. This iontronic touch panel with skin-friendly and wearable qualitities is helpful for next-generation wearable interaction electronics.

Artificial Multisensory Neuron with a Single Transistor for Multimodal Perception through Hybrid Visual and Thermal Sensing

An artificial multisensory device applicable to in-sensor computing is demonstrated with a single-transistor neuron (1T-neuron) for multimodal perception. It simultaneously receives two sensing signals from visual and thermal stimuli. The 1T-neuron transforms these signals into electrical signals in the form of spiking and then fires them for a spiking neural network at the same time. This feature makes it feasible to realize input neurons for multimodal sensing. Visual and thermal sensing is achieved due to the inherent optical and thermal behaviors of the 1T-neuron. To demonstrate a neuromorphic multimodal sensing system with the artificial multisensory 1T-neuron, fingerprint recognition, widely used for biometric security, is implemented. Owing to the simultaneous sensing of heat as well as light, the proposed fingerprint recognition system composed of multisensory 1T-neurons not only identifies a genuine pattern but also judges whether or not it is forged.