Abstract The rapidly increasing demand for data processing at user interfaces underscores the critical need for efficient time‐series data handling in wearable and portable electronics. Reservoir computing (RC), which emulates biological computation, holds significant promise for processing temporal information. It comprises physical reservoirs and readout layers that transform time‐series signals and enable parallel computation, respectively. However, full integration of hardware RC systems on a single flexible substrate remains challenging due to the distinct functional requirements of reservoir and readout nodes. Here, a small‐molecule‐based memristive framework is presented tailored for flexible RC systems. Flexible memristor arrays incorporating a multifunctional interfacial layer that enables tunable grain distributions in the switching layer exhibit biologically inspired short‐ and long‐term plasticity, key for implementing physical reservoirs and readout networks, in a grain‐size‐dependent manner. In addition, the memristor arrays exhibit high reliability and uniform performance, with low device‐to‐device and cycle‐to‐cycle variations (≈7.87% and ≈5.19%, respectively). RC systems based on this framework exhibit efficient data compression and robust adaptability to temporal variations, such as rotational transformations, in handwritten digit recognition. These small‐molecule memristive platforms provide a promising hardware foundation for intelligent, flexible, and energy‐efficient wearable electronics.

Flexible neuromorphic architectures that emulate biological cognitive systems hold great promise for smart wearable electronics. To realize neuro-inspired sensing and computing electronics, artificial sensory neurons that detect and process external stimuli must be integrated with central nervous systems capable of parallel computation. In near-sensor computing, synaptic devices, and sensors are used to emulate sensory neurons and receptors, respectively. In contrast, in in-sensor computing, a single multifunctional device serves as both the receptor and neuron. Bio-inspired cognitive systems efficiently detect and process stimuli through data structuring techniques, significantly reducing data volume and enabling the extension of neuromorphic applications to smart wearable systems. To construct wearable near- and in-sensor computing, it is crucial to develop artificial sensory neurons and central nervous synapses that replicate the biological functionalities. Additionally, the integrated systems must exhibit high mechanical flexibility and integration density. This review addresses research on flexible bio-inspired cognitive systems, classified into near- and in-sensor computing. It covers fundamental aspects, including biological cognitive processes, the required components, and the structures for each component, as well as applications for wearable smart systems. Finally, it offers perspectives on future research directions for flexible neuromorphic electronics in smart wearable systems connected to the next-generation Internet of Things.

Abstract The rapidly increasing demand for data processing at user interfaces underscores the critical need for efficient time‐series data handling in wearable and portable electronics. Reservoir computing (RC), which emulates biological computation, holds significant promise for processing temporal information. It comprises physical reservoirs and readout layers that transform time‐series signals and enable parallel computation, respectively. However, full integration of hardware RC systems on a single flexible substrate remains challenging due to the distinct functional requirements of reservoir and readout nodes. Here, a small‐molecule‐based memristive framework is presented tailored for flexible RC systems. Flexible memristor arrays incorporating a multifunctional interfacial layer that enables tunable grain distributions in the switching layer exhibit biologically inspired short‐ and long‐term plasticity, key for implementing physical reservoirs and readout networks, in a grain‐size‐dependent manner. In addition, the memristor arrays exhibit high reliability and uniform performance, with low device‐to‐device and cycle‐to‐cycle variations (≈7.87% and ≈5.19%, respectively). RC systems based on this framework exhibit efficient data compression and robust adaptability to temporal variations, such as rotational transformations, in handwritten digit recognition. These small‐molecule memristive platforms provide a promising hardware foundation for intelligent, flexible, and energy‐efficient wearable electronics.

Flexible neuromorphic architectures that emulate biological cognitive systems hold great promise for smart wearable electronics. To realize neuro-inspired sensing and computing electronics, artificial sensory neurons that detect and process external stimuli must be integrated with central nervous systems capable of parallel computation. In near-sensor computing, synaptic devices, and sensors are used to emulate sensory neurons and receptors, respectively. In contrast, in in-sensor computing, a single multifunctional device serves as both the receptor and neuron. Bio-inspired cognitive systems efficiently detect and process stimuli through data structuring techniques, significantly reducing data volume and enabling the extension of neuromorphic applications to smart wearable systems. To construct wearable near- and in-sensor computing, it is crucial to develop artificial sensory neurons and central nervous synapses that replicate the biological functionalities. Additionally, the integrated systems must exhibit high mechanical flexibility and integration density. This review addresses research on flexible bio-inspired cognitive systems, classified into near- and in-sensor computing. It covers fundamental aspects, including biological cognitive processes, the required components, and the structures for each component, as well as applications for wearable smart systems. Finally, it offers perspectives on future research directions for flexible neuromorphic electronics in smart wearable systems connected to the next-generation Internet of Things.

The increasing menace of counterfeiting and information theft underscores the urgent need for security platforms compatible with both micro- and nanoelectronics. Existing methods for anticounterfeiting labeling and cryptographic systems rely on unclonable patterns derived from the unpredictable variability of physical phenomena. However, these approaches impose limitations on the scalability of security components. Here we present a scalable platform for photoresponsive physically unclonable functions based on oxide particle kinetics in polymer solutions. The stochastic agglomeration process occurring during the formation of polymer films with dispersed oxide particles yields random patterns, with pixel sizes scalable from micro to nanoscales. We produce mechanically flexible and self-destructible optical unclonable function patterns utilizing oxide aggregates on a polymer film. Moreover, we establish a strategy for generating electrical unclonable patterns on a conducting polymer film. This involves covering the polymer film with an aggregate pattern mask, which serves as an ultraviolet-blocking layer for randomly exposing the film to ultraviolet ozone treatment. These unclonable patterns constitute robust and compact security systems, exhibiting effective resilience against machine-learning attacks (∼50% prediction error for training data sets of 1000). The developed scalable platforms for physically unclonable functions provide a hardware solution for robust cryptographic applications.

Dual-gate multi-bit memory transistors that combine charge-trap and ferroelectric effects are demonstrated as a synaptic memory with high dynamic range.

Despite the advantages of solution-processed organic ferroelectric transistors as synaptic components, such as stable memory states and fast switching speeds, the realization of key synaptic functions, including continuous weight modulation and low energy consumption, remains challenging. In this study, we present a strategy to optimize the charge injection barrier at the source-semiconductor interface to enhance synaptic functionalities. By incorporating heterobimetallic electrodes, we systematically tailor the hole injection barrier to suppress leakage current in the memory-off state while inducing thermionic emission-dominated channel conduction in the memory-on state. This approach enables low operating currents and facilitates the gradual modulation of channel conductance. The optimized devices exhibit a high memory on/off ratio (∼10⁴) with low off-state currents, as well as linearly tunable memory states with a low nonlinearity factor (∼1.68), making them suitable for practical hardware neural networks. Owing to these improved synaptic properties, hardware neural networks incorporating these devices demonstrate high recognition accuracy in handwritten digit classification tasks. This approach lays a foundation for the development of portable and flexible neuromorphic systems, approaching biological levels of functionality.

Brain-inspired neuromorphic electronics have been extensively studied as systems for wearable devices, neuroprostheses, and soft machines, offering solutions to the limitations of conventional von Neumann computing systems and enabling efficient information processing. Among these, synaptic transistors with vertical structures are gaining significant attention as promising candidates for flexible neuromorphic electronics, owing to their unique structural features, such as ultrashort channel lengths and vertical carrier transport, which provide superior performance, mechanical flexibility, and high-density integration. Vertical synaptic transistors (VSTs) not only combine the functionalities of information processing, memory, and sensing/responding within a single device but also enable the realization of diverse synaptic properties, effectively mimicking the information processing and sensory capabilities of biological nervous systems. Achieving both mechanical flexibility and excellent electrical performance in VSTs necessitates a strong focus on the active layer, prompting extensive research into various flexible semiconducting materials. This review explores the diverse range of flexible semiconducting materials employed in VSTs and their fundamental operating mechanisms. Additionally, it highlights recent advancements in VSTs and systems developed to replicate the functionalities of biological nervous systems.

The basic structure of resistive random access memory (RRAM), with an insulator between two metal electrodes, closely resembles that of a capacitor. However, most studies have focused on resistive switching characteristics, with little attention to the coexistence with capacitive switching. In this study, we analyzed the coexistence of resistance and capacitance memory effects in RRAM devices through measurements and simulations. Using an Al/Al₂O₃/HfO₂/SiO_x/p⁺-Si stack, we confirmed this coexistence experimentally. DC measurements showed bipolar switching characteristics with set operations at positive voltages. Continuous set and reset pulse measurements revealed that the low resistance state (LRS) coincides with a high capacitance state (HCS), while the high resistance state (HRS) aligns with a low capacitance state (LCS). To investigate the underlying mechanisms, we conducted two simulations using COMSOL Multiphysics to confirm the overall trend of device memory effects: One simulation focused on oxide-silicon interface trap charge variations that induce capacitance changes due to Joule heating and heat transfer, while another simulation focused on the capacitance variations induced by changes in the oxygen ion concentration within the oxygen reservoir layer according to the device's resistance state.

Abstract Memristors are promising candidates for generating physically unclonable functions (PUFs) due to their inherently stochastic resistive switching behaviors. In memristive devices, conducting paths are formed via charge trapping, ion migration, or filamentary conduction mechanisms across the active region, resulting in arbitrary resistive switching characteristics. To harness memristors for PUF applications, it is essential to enhance the entropy sources associated with device operation to optimize randomness, uniqueness, and diffuseness. In this review, we elucidate the fundamental mechanisms underlying resistive switching in memristors, with a focus on factors contributing to their stochastic characteristics. Recent strategies to enhance randomness, ranging from material and device-level engineering to external circuit algorithms, are also discussed. Furthermore, practical applications of memristor-based PUFs, including cryptographic key generation, data encryption, and random data generation, are introduced. Finally, we outline future research directions for advancing memristor-based PUFs toward widespread deployment across diverse application domains.

Abstract Neuromorphic motor control systems aim to emulate the adaptive and efficient motor regulation observed in biological organisms by seamlessly integrating sensing and processing with actuation. Electrolyte‐gated organic synaptic transistors (EGOSTs) have emerged as promising building blocks for such systems due to their ability to mimic synaptic behavior through ion–electronic coupling, analogous to biological synapses. This review highlights recent advances in EGOST‐based neuromorphic motor control systems, focusing on their operational mechanisms, biological synaptic plasticity characteristics, and integration with motor actuators. A biological perspective on motor control is provided, emphasizing the roles of synaptic transmission and plasticity. It is then examined how these functions are emulated in EGOSTs, including strategies for tuning device behavior through morphology control and incorporating intrinsic sensing capabilities within a single device. Applications are categorized across artificial muscle fibers, robotic manipulators, and neuromuscular prostheses, demonstrating the versatility of EGOSTs in enabling low‐power, adaptive, and biointegrated motion control. Finally, key challenges—such as material limitations, electrochemical stability, and system‐level integration are discussed—that must be addressed to transition from proof‐of‐concept demonstrations to real‐world applications. This review underscores the transformative potential of EGOST‐based neuromorphic platforms for future wearable robotics, neuroprosthetics, and bioinspired intelligent systems.