Search

This letter introduces ERRA, an embodied learning architecture that enables robots to jointly obtain three fundamental capabilities (reasoning, planning, and interaction) for solving long-horizon language-conditioned manipulation tasks. ERRA is based on tightly-coupled probabilistic inferences at two granularity levels. Coarse-resolution inference is formulated as sequence generation through a large language model, which infers action language from natural language instruction and environment state. The robot then zooms to the fine-resolution inference part to perform the concrete action corresponding to the action language. Fine-resolution inference is constructed as a Markov decision process, which takes action language and environmental sensing as observations and outputs the action. The results of action execution in environments provide feedback for subsequent coarse-resolution reasoning. Such coarse-to-fine inference allows the robot to decompose and achieve long-horizon tasks interactively. In extensive experiments, we show that ERRA can complete various long-horizon manipulation tasks specified by abstract language instructions. We also demonstrate successful generalization to the novel but similar natural language instructions., Comment: Accepted to IEEE Robotics and Automation Letters (RA-L)

Humans excel in grasping objects through diverse and robust policies, many of which are so probabilistically rare that exploration-based learning methods hardly observe and learn. Inspired by the human learning process, we propose a method to extract and exploit latent intents from demonstrations, and then learn diverse and robust grasping policies through self-exploration. The resulting policy can grasp challenging objects in various environments with an off-the-shelf parallel gripper. The key component is a learned intention estimator, which maps gripper pose and visual sensory to a set of sub-intents covering important phases of the grasping movement. Sub-intents can be used to build an intrinsic reward to guide policy learning. The learned policy demonstrates remarkable zero-shot generalization from simulation to the real world while retaining its robustness against states that have never been encountered during training, novel objects such as protractors and user manuals, and environments such as the cluttered conveyor., Comment: Accepted to IEEE Robotics and Automation Letters (RA-L)

AZO@Ag nanowire@AZO (AAA) composite stretchable transparent electrode enables full-solution-processed semi-transparent intrinsically stretchable OPV achieving record 10.90% PCE, with excellent mechanical robustness against stretch-release cycles.

The rapid growth in the miniaturized mechanical and electronic devices industry has created the need for temporary attachment systems that can carry out pick-and-place and transfer printing tasks for fragile and tiny parts. Current systems are limited by a fundamental trade-off between adhesive strength and state-changing trigger force, which causes the need for a rapidly switchable adhesive. In this study, an elastomeric microstructure is presented combining a trapezoidal-prism-shaped (TPS) and a mushroom-shaped microstructure, which overcomes the trade-off with the help of the TPS structure. The optimal design exhibits a strong adhesive strength of 87.8 kPa and a negligible detachment strength of, Advanced Functional Materials, ISSN:1616-301X, ISSN:1616-3028

Earthquake stress drop values estimated from ground-motion spectra commonly vary by several orders of magnitude, particularly for small earthquakes (~M < 3). Stress-drop values have been found to vary with faulting style, faulting type (intraplate, interplate), depth, and to exhibit differences between natural and induced earthquakes. Nevertheless, distinguishing uncertainties from real trends across data sets is challenging, in part due to the variation in methodological approaches and observational constraints. However, the proliferation of high-quality, dense seismic data in recent years has shown that at least some of the variability in stress drop values almost certainly reflects diversity in fault strength and geological conditions. Coupling well-constrained observations to a variety of modeling approaches will help uncover what controls earthquake rupture processes, but deconvolving observational constraints from real variation in rupture behavior is key.We present our stress drop estimates from data sets representing a wide range of fault loading conditions and geological environments, from interplate, intraslab and forearc subduction faults, to volcanic, intraplate, and human induced events. Stress-drop values range primarily between 1 – 100 MPa for events that meet the criteria for spectral-ratio analysis. We present correlations of low relative stress drop values in areas of high seismic attenuation indicative of lower rock strength, and a slight correlation with depth that corresponds to modeled deviatoric stress values. We also show one notable subset of induced events near active injection wells that exhibit stress drop values of ~0.1 MPa and have distinctive low-frequency content. Their spatial distribution, waveform, and source spectral characteristics suggest either slower rupture, lower stress drop values, or a combination of both, and may represent part of the transition between aseismic and seismic slip. We show using a Large-n array that while stress drop values are roughly constant (within 2 orders of magnitude), estimates can vary by roughly 25% when station coverage is limited to 15 stations or less with a maximum azimuthal gap of 90°. Our findings highlight the importance of using modeling approaches to explore relative influence of fault strength and methodological approaches in stress drop variation. In particular, models that incorporate both frictional and thermoelastic approaches may provide clues to the variability of conditions that can activate faults, both within stable sliding and seismic rupture conditions.

The development of machine learning interatomic potentials has immensely contributed to the accuracy of simulations of molecules and crystals. However, creating interatomic potentials for magnetic systems that account for both magnetic moments and structural degrees of freedom remains a challenge. This work introduces SpinGNN, a spin-dependent interatomic potential approach that employs the graph neural network (GNN) to describe magnetic systems. SpinGNN consists of two types of edge GNNs: Heisenberg edge GNN (HEGNN) and spin-distance edge GNN (SEGNN). HEGNN is tailored to capture Heisenberg-type spin-lattice interactions, while SEGNN accurately models multi-body and high-order spin-lattice coupling. The effectiveness of SpinGNN is demonstrated by its exceptional precision in fitting a high-order spin Hamiltonian and two complex spin-lattice Hamiltonians with great precision. Furthermore, it successfully models the subtle spin-lattice coupling in BiFeO3 and performs large-scale spin-lattice dynamics simulations, predicting its antiferromagnetic ground state, magnetic phase transition, and domain wall energy landscape with high accuracy. Our study broadens the scope of graph neural network potentials to magnetic systems, serving as a foundation for carrying out large-scale spin-lattice dynamic simulations of such systems., Comment: 28 pages, 4 figures

Robotic grippers with visuotactile sensors have access to rich tactile information for grasping tasks but encounter difficulty in partially encompassing large objects with sufficient grip force. While hierarchical gecko-inspired adhesives are a potential technique for bridging performance gaps, they require a large contact area for efficient usage. In this work, we present a new version of an adaptive gecko gripper called Viko 2.0 that effectively combines the advantage of adhesives and visuotactile sensors. Compared with a non-hierarchical structure, a hierarchical structure with a multimaterial design achieves approximately a 1.5 times increase in normal adhesion and double in contact area. The integrated visuotactile sensor captures a deformation image of the hierarchical structure and provides a real-time measurement of contact area, shear force, and incipient slip detection at 24 Hz. The gripper is implemented on a robotic arm to demonstrate an adaptive grasping pose based on contact area, and grasps objects with a wide range of geometries and textures., This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

Recent studies confirm that the distribution of injection-induced earthquakes (IIE) can be related to both natural (e.g., tectonic, geological, and hydrological) settings and operational details. However, the relative importance of operational factors with respect to the natural ones has not been fully understood for the western Canada sedimentary basin. In this study, we train the eXtreme Gradient Boosting (XGBoost) machine-learning algorithm to comprehensively evaluate six geological and seven industrial operational factors suspected to be correlated with the distribution of IIE in the northern Montney play (NMP), British Columbia. We then derive the Shapley Additive Explanations values to quantitatively interpret the outputs from XGBoost. Our results reveal that operational and geological factors have comparable contributions to the IIE distribution. The top four features that contribute most to the seismicity pattern are horizontal distance to the Cordilleran deformation front, cumulative injected volume, shut-in pressure and vertical distance to the Debolt formation (with respect to the hydraulic fracturing [HF] depth). Features with secondary influence are the thickness of the Montney formation, breakdown pressure, cumulative fault length per unit area, and vertical distance to the basement (with respect to the HF depth). Other remaining features (e.g., the average treating pressure and injection rate) appear the least related. Our results provide critical information to establishing a comprehensive susceptibility model that includes key geological and operational factors affecting the IIE distribution in the NMP area.

Background Apatinib (YN968D1) is the first small-molecule-targeting drug with anti-tumor activity created in China for the treatment of advanced gastric cancer (GC) and hepatocellular carcinoma (HCC). It showed significant variation in the efficacy for treating cancers, including advanced non-squamous non-small-cell lung cancer (NSCLC). Whether its efficacy could be optimized by subgrouping patients with certain genetic variation remains elusive. Methods Here, we firstly used kinase screening to identify any possible target of apatinib against 138 kinases. The effects of apatinib on proliferation rates, cell cycle, cell apoptosis, and cell migration on cancer cell lines were analyzed; the in vitro potential pathways of apatinib on cancer cell lines were screened. The effect of apatinib on mouse cancer models in vivo was also analyzed. Results Based on HCC364 cells with BRAF V600E mutation, we have shown that apatinib could inhibit their growth, migration, cell cycle, and induce their apoptosis. Based on mice with transplanted HCC364 cells, we have also shown that apatinib could inhibit the tumor growth. Based on immunohistochemistry, we have demonstrated that apatinib could suppress the phosphorylation of mitogen-activated protein kinase/extracellular signal-regulated kinase and extracellular regulated protein kinases. This may account at least part of the apatinib’s inhibitory effect on HCC364 cancer cells. Conclusions BRAF V600E protein kinase is a target of apatinib by kinase screening. We have demonstrated that apatinib can effectively inhibit tumor cells with BRAF V600E mutation by in vitro and in vivo experiments. Our results have demonstrated that targeting BRAF V600E mutation, apatinib appears to be effective and safe for treating NSCLC and possibly other cancers with the same mutation.

In this work, a stretchable thin film capacitive strain sensor based on wavy-structured interdigitated metal electrodes is successfully demonstrated. Wavy-shape metal electrodes were patterned using bulk silicon micromachining, endowing the electrodes with high stretchability and reliability. The electrodes were also protected with a layer of parylene C to prevent damage during stretching and embedded in polydimethylsiloxane (PDMS) stretchable package layers. The interdigitated electrodes sitting on the wavy structures experienced a waving angle change under external strain resulting in capacitive change, which is the sensing mechanism of the wavy structured interdigitated capacitive strain sensor (WICSS). The sensitivity and stretchability of the WICSS is dependent on the design of interdigitated electrodes and the dimension of the wavy structures. The gauge factor (GF) of the WICSS was 0.27 at 25% strain. There was little hysteresis during the stretching and releasing process in the range of 25% strain. The sensing property remained stable within a 1000 cycling test. The applications of the high-performance WICSS are also demonstrated by detecting the motion of human fingers and wrists.

We present a vertically aligned multiwall carbon nanotube (VA-MWCNT) based humidity sensor with low cost and simple fabrication method compatible with mass production. The VA-MWCNT is well aligned and possesses an interconnect topography. Initially, the VA-MWCNT showed superhydrophobic characteristics, which are not suitable for humidity sensing. Therefore, to protect the topography of the material and make the surface more hydrophilic, a low cost and easy handling treatment method was adopted with O2 plasma surface functionalization. The fabricated device showed good linearity at the range of 40% RH–90% RH, low hysteresis ( H = 1.5%RH), fast response (3.3 s), high stability ( Cv Rr _max 0.6% ), demonstrating excellent performance in both static and dynamic measurements. The high performance and low-cost fabrication methods make the VA-MWCNT-based sensor a promising candidate for the application of humidity sensing. Moreover, benefiting from the simple fabrication process, it can be easily extended to resistive sensing fusion with the integration of the temperature sensor and flow sensor for more complex applications.

BackgroundIn China, some patients avoid seeking medical care and are highly sensitive to subsequent medical care because of fear of possible hospitalization after a diagnosis has been established. Early identification of fear of hospitalization is essential for clinical staff to develop targeted education and interventions. However, there are currently no tools to assess outpatients’ fear of hospitalization in mainland China. This study aimed to translate the Fear of Hospitalization (FH) scale into Chinese and verify its reliability and validity in outpatients.MethodsThrough convenience sampling, 664 outpatients who required hospitalization were recruited from two cities in Liaoning Province, China. The reliability of the translated scale was measured by internal consistency, split-half reliability, and test–retest reliability. The validity of the translated scale was evaluated by expert consultation, exploratory factor analysis, and confirmatory factor analysis. Data were analyzed using SPSS 25.0 (IBM Corp., Armonk, NY, United States) and AMOS 23.0 (IBM Corp., Armonk, NY, United States).ResultsThe Cronbach’s α value of the Chinese version of the FH scale was 0.849, and the Cronbach’s α value of the dimensions ranged from 0.857 to 0.902. The test–retest reliability value of 0.868 shows good temporal stability. The split-half reliability value of 0.910 indicates a high degree of measuring the same content. The content validity index of the scale (S-CVI) was 0.924, indicating a good level of content validity. The 3-factor structure supported by eigenvalues, total variance explained, and scree plot was obtained using exploratory factor analysis. In addition, all recommended fit indicators were within the acceptable range by confirmatory factor analysis.ConclusionThe Chinese version of the FH scale is valid and reliable in outpatients. The developed three-factor structured scale will help identify outpatients with a high fear of hospitalization and can inform the development of educational intervention plans for care managers, physicians, and nurses. In addition, it helps clinicians and nurses take action to reduce this fear of hospitalization in patients and prevent avoidance of using health care services due to fear of hospitalization.

Multiplexed DNA fluorescence in situ hybridization (FISH) imaging technologies have been developed to map the folding of chromatin fibers at tens of nanometer and tens of kilobase resolution in single cells. However, computational methods to reliably identify chromatin loops from such imaging datasets are still lacking. Here we present a Single-Nucleus Analysis Pipeline for multiplexed DNA FISH (SnapFISH), to process the multiplexed DNA FISH data and identify chromatin loops. SnapFISH can identify known chromatin loops from mouse embryonic stem cells with high sensitivity and accuracy. In addition, SnapFISH obtained comparable results of chromatin loops across datasets generated from diverse imaging technologies.

This work presents Time-reversal Equivariant Neural Network (TENN) framework. With TENN, the time-reversal symmetry is considered in the equivariant neural network (ENN), which generalizes the ENN to consider physical quantities related to time-reversal symmetry such as spin and velocity of atoms. TENN-e3, as the time-reversal-extension of E(3) equivariant neural network, is developed to keep the Time-reversal E(3) equivariant with consideration of whether to include the spin-orbit effect for both collinear and non-collinear magnetic moments situations for magnetic material. TENN-e3 can construct spin neural network potential and the Hamiltonian of magnetic material from ab-initio calculations. Time-reversal-E(3)-equivariant convolutions for interactions of spinor and geometric tensors are employed in TENN-e3. Compared to the popular ENN, TENN-e3 can describe the complex spin-lattice coupling with high accuracy and keep time-reversal symmetry which is not preserved in the existing E(3)-equivariant model. Also, the Hamiltonian of magnetic material with time-reversal symmetry can be built with TENN-e3. TENN paves a new way to spin-lattice dynamics simulations over long-time scales and electronic structure calculations of large-scale magnetic materials., Comment: 17 pages,2 figures and 2 tables

Background: Hepatic fibrosis (HF) is characterized by activation of hepatic stellate cells (HSCs) and extensive deposition of extracellular matrix components, especially collagens. However, effective antifibrotic therapies are still lacking. Recently, circular RNAs (circRNAs) have been identified as novel regulators of HF.Methods: circRNAs profile was screened by RNA sequencing and the location of circ_0008494 was confirmed by fluorescence in situ hybridization assay in human HF tissues. Bioinformatics analysis was used for result prediction and dual-luciferase reporter, together with AGO-RIP and biotin-coupled miRNA capture assays, were used to determine miR-185-3p/collagen type I alpha 1 chain (Col1a1) as the target of circ_0008494. A stable circ_0008494-interfering human HSCs cell line was constructed and used to determine the regulatory mechanism of circ_0008494/miR-185-3p/Col1a1 axis.Results: circ_0008494 was abundantly and significantly over-expressed in human HF tissues and located at the cytoplasm of HSCs. Together, dual-luciferase reporter, AGO-RIP and biotin-coupled miRNA capture assays confirmed that circ_0008494 acted as a sponge of miR-185-3p. Cell functional experiments and rescue assays demonstrated suppressing circ_0008494 could inhibit activation, proliferation, migration of HSCs and promote their apoptosis through miR-185-3p. In particular, the HF indicator, Col1a1, was validated as the direct target of miR-185-3p and the suppression of circ_0008494 inhibited the expression of Col1a1 by releasing miR-185-3p.Conclusion: Knocking down circ_0008494 inhibited HSCs activation through the miR-185-3p/Col1a1 axis. circ_0008494 could be a promising treatment target for HF.

The influence of doping on the high-Tc superconductor Im-3 m-H3S has attracted great attention since the addition of the external elements is expected to further increase the Tc value or lower the stable pressure of H3S. Here, we systematically investigated the electronic structures and superconductive properties of H6SX (X = N, P, As, O, Se, F, Cl, and Br) through first-principle calculations at 200 GPa. Only H6SP, H6SSe, and H6SBr are dynamically stable and the Tc of H6SSe can reach 182 K. The bonding analysis shows that stronger covalent interactions are beneficial for higher Tc values for these ternary metallic covalent superconductors. The improvement of Tc in doped H3S systems could be obtained through introducing stronger covalent interactions and light atomic massive. In addition, we found that the replacement of Br atoms can reduce the stable pressure of H3S to 150 GPa with Tc about 159–174 K.

In this work, we report on the fabrication and characterization of a miniature GaN chip with a size of $1.1\times 1.1\times0.21$ mm3 for surface roughness measurement. Unlike conventional optical metrology that requires external components for the optical coupling between light source and detector, a monolithic GaN-on-sapphire chip fabricated through wafer-scale processes provides light emission and photodetection functions and is capable of sensing a range of average surface roughness from 0.025 to $1.6~\mu \text{m}$ . The sensing performances are compared with fiber-optic metrology, and the ability to real-time monitor the roughness of the moving surface is demonstrated. The novel chip offers the advantage of non-contact and non-destructive, compact structure, simple operation, and in situ measurement capability, which is highly valuable for various practical and industrial surface metrology applications.

This paper presented the principle, design, fabrication, and characterization of Molecular Electronic Transducer (MET) single-axis tilt sensors. The proposed sensor has a circular channel deposited with symmetric concentric electrode pairs and half-filled with a liquid electrolyte. Such inclination-dependent liquid level is detected by the differential current from the electrochemical reaction at the two symmetric concentric electrode pairs. In order to reduce the form factor, an organic solvent is added to the aqueous electrolyte to lower the surface tension force. Limited by the liquid motion and reaction rate, the sensors are capable to measure tilt angle under static conditions only. From the tilt angle stepping experiments, the performance of the proposed tilt sensors is characterized and compared. The sensors’ response time is dominated by the electrochemical equilibrium arrival time. They can achieve a low discrimination threshold of 0.016° and a short time constant of 0584s within the ±85° measurement range with a power consumption of less than 0.1mW. Compared to previously reported MET tilt sensors with a parallel swirl electrode configuration, the signal-to-noise ratio and response time are improved significantly.

Motion sensor is a kind of sensor which is commonly used in the field of human-computer interaction. This article uses Microsoft’s Kinect sensor to get in-depth information about gesture recognition in electronic instructional demonstrations. Here, we present the Kinect V2 sensor for use in demonstrating gesture capture and recognition in electronic information instruction. In this study, depth information is used to dynamically capture electronic information teaching demonstrative gestures. In addition, a combined denoising method is also proposed, which can effectively remove the interference and noise compared with the single denoising method. Then, the researchers programmed and moved the dynamic recognition of demonstrative gestures in electronic information instruction. Compared with some traditional denoising methods, the combined denoising method can effectively remove the blur and boundary burr. This study can be applied to the field of human-computer interaction in electronic information teaching to further improve the accuracy of information.

Lacking of S-wave velocity in logging will negatively affect the process of reservoir prediction and formation evaluation for organic rich rock reservoir. In this paper, a simple but effective method is presented to predict the S-wave velocity from P-wave velocity based on the combination of the critical porosity consolidation coefficient (CPCC) model with dual-variable parameters and quantum particle swarm optimization (QPSO) algorithm. In the presented method, the P-and S-wave velocities of organic-rich rock are linked by consolidation coefficient and critical porosity. Then, the error between the measured and calculated P-wave velocity is used to establish the inverse objective function. Moreover, the QPSO algorithm is introduced to invert the critical porosity and consolidation coefficient. Finally, the S-wave velocity can be predicted using the inverted parameters. Compared with two existing single-parameter adaptive methods, the predicted S-wave velocity of the proposed method is superior, which can be satisfactorily implemented for laboratory measurement and on logging data.

BackgroundIn China, clinical deterioration events present a real problem for every clinical nurse. Patient deterioration is determined in part by nurses’ attitudes toward early recognition of clinical deterioration. However, research on attitudes toward the early identification of clinical deterioration is still in its infancy, and even less research has been done on ward nurses’ attitudes toward the early identification of clinical deterioration. To drive behavioral change and improve the care of deteriorating patients, nurses need comprehensive, valid, and reliable tools to assess their attitudes toward early identification of deterioration.ObjectiveIn this study, we aimed to translate the Attitudes Toward Recognizing Early and Noticeable Deterioration (ATREND) scale into Chinese and to assess its validity and reliability tests.MethodsFrom March 2022 to July 2022, the ATREND scale was translated, back-translated, and cross-culturally adapted into the Chinese version using a modified Brislin translation model. Then, 460 ward nurses were recruited from tertiary Grade A general hospitals in two cities: Shenyang and Jinzhou in Liaoning Province, China. Reliability analyses were conducted using internal consistency, split-half, and test–retest reliability. We convened a committee of experts to determine the validity of the content. Tests of the structural validity of the scale were conducted using exploratory and validation factor analyses.ResultsThe Cronbach’s α value of the Chinese version of the ATREND scale was 0.804, and the Cronbach’s α value of the dimensions ranged from 0.782 to 0.863. The split-half reliability and test–retest reliability were 0.846 and 0.711, respectively. Furthermore, the scale has an index of content validity of 0.922, indicating a high level of content validity. In exploratory factor analysis, eigenvalues, total variance explained, and scree plot supported a three-factor structure. The three-factor model supported by this study was confirmed by confirmatory factor analysis (CFA). Moreover, the model fitting indexes (e.g., χ2/DF = 1.498, GFI = 0.954, RMSEA = 0.047) were all within acceptable limits based on the CFA.ConclusionThe Chinese version of the scale is reliable and valid among ward nurses. Nursing educators and clinicians will be able to develop targeted educational programs to enhance the competence and behaviors of Chinese ward nurses in recognizing clinical deterioration. It will be based on the developed scale to assess Chinese nurses’ attitudes and practices regarding early recognition of clinical deterioration. As a result, it is necessary to consider the Chinese scale’s three-factor structure. The developed three-factor structured scale will assess Chinese ward nurses’ attitudes and practices toward patient observation and vital sign-monitoring empowerment, enlightening them on the importance of patient observation, encouraging ward nurses to use a wider range of patient assessment techniques to capture early signs of clinical deterioration, and helping ward nurses to develop clinical confidence to monitor clinical deterioration.

Single cell Hi-C (scHi-C) has been used to map genome organization in complex tissues. However, computational tools to detect dynamic chromatin contacts from scHi-C datasets in development and through disease pathogenesis are still lacking. Here, we present SnapHiC-D, a computational pipeline to identify differential chromatin contacts (DCCs) between two scHi-C datasets. Compared to methods designed for bulk Hi-C data, SnapHiC-D detects DCCs with high sensitivity and accuracy. We used SnapHiC-D to identify celltype-specific chromatin contacts at 10 kilobase resolution in mouse hippocampal and human prefrontal cortical tissues, and demonstrated that DCCs detected in the cortical and hippocampal cell types are generally correlated with cell-type-specific gene expression patterns and epigenomic features.

Graph neural networks (GNNs) have shown extremely high flexibility and accuracy in predicting the physical properties of molecules and crystals. However, traditional invariant GNNs are not compatible with directional properties, which currently limits their application only to the prediction of invariant scalar properties. To address this issue, here we propose a general framework, i.e., an edge-based tensor prediction graph neural network (ETGNN), where a tensor is expressed as the linear combination of the local spatial components projected on the edge directions of multiple-sized clusters. This tensor decomposition is rotationally equivariant and exactly satisfies the symmetry of the local structures, while the coefficients before each local spatial component are rotationally invariant scalars which can be easily predicted by traditional invariant GNNs. The accuracy and universality of our new framework are demonstrated by the successful prediction of various tensor properties from first-order to third-order. The framework proposed in this work will enable GNNs to step into the broad field of the prediction of directional properties.

Flexible and stretchable humidity sensors for wearable purposes have become increasingly important in health care and physiological signal monitoring. However, to the authors' knowledge, there is no report on flexible and stretchable paper-based humidity sensors that are low-cost, easily fabricated, and environmentally friendly. In this work, for the first time, we propose a stretchable, textile-compatible paper-based origami humidity sensor (POHS). The POHS can achieve good stretchability by integrating origami folding structures with a paper substrate, in which an airlaid paper acts as both a sensing material and a sensor substrate. This sensor has high sensitivity, good response, and recovery properties with excellent stability during deformation. This sensor has proved to be capable of dynamically monitoring the breathing rate after 300 folding and unfolding cycles. The flexible and stretchable nature of our POHS ensures that it is compatible for textile attachment and its utility for wearable applications, including respiration rate monitoring and diaper wetting detection. The facile fabrication process and convenient disposal method of the POHS proposed in this study provide feasible solutions for the development of low-cost wearable humidity sensors.