Environmental changes and sensor aging can cause sensor drift in sensor array responses (i.e., a shift in the measured signal/feature distribution over time), which in turn degrades gas classification performance in real-world deployments of electronic-nose systems. Previous studies using the UCI Gas Sensor Array Drift Dataset as a benchmark reported promising drift compensation results but often lacked robust statistical validation and may overcompensate for drift by suppressing class-discriminative variance. To address these limitations and rigorously evaluate improvements in sensor-drift compensation, we designed two domain adaptation tasks based on the UCI electronic-nose dataset: (1) using the first batch to predict remaining batches, simulating a controlled laboratory setting, and (2) using Batches 1 through n−1 to predict Batch n, simulating continuous training data updates for online training. Then, we systematically tested three methods—our semi-supervised knowledge distillation method (KD) for sensor-drift compensation; a previously benchmarked method, Domain-Regularized Component Analysis (DRCA); and a hybrid method, KD–DRCA—across 30 random test-set partitions on the UCI dataset. We showed that semi-supervised KD consistently outperformed both DRCA and KD–DRCA, achieving up to 18% and 15% relative improvements in accuracy and F1-score, respectively, over the baseline, proving KD’s superior effectiveness in electronic-nose drift compensation. This work provides a rigorous statistical validation of KD for electronic-nose drift compensation under long-term temporal drift, with repeated randomized evaluation and significance testing, and demonstrates consistent improvements over DRCA on the UCI drift benchmark.

Abstract When charge transport occurs under conditions such as topological protection or ballistic motion, the conductance of low-dimensional systems often exhibits quantized values in units of e 2 / h , underpinning advances in quantum metrology and computing. Here we report a quantized quantity: the ratio of displacement field to magnetic field, D / B , in large-twist-angle bilayer graphene. In high magnetic fields, Landau-level crossings between the top and bottom layers produce equal-sized checkerboard patterns across the D / B – ν space. These arise from electric-field-driven interlayer charge transfer of one elementary charge per flux quantum, yielding quantized critical displacement intervals, δ D = $$\frac{e}{2{{\uppi }}{l}_{B}^{2}}$$ e 2 π l B 2 , where l B is the magnetic length. This mechanism offers a route to magnetic sensing, as the displacement-to-magnetic-field ratio is defined solely by fundamental constants. We propose a prototype magnetometer based on this principle, potentially enabling planar mapping of magnetic fields with micrometre resolution via large-twist-angle bilayer graphene sensor arrays. Our results demonstrate that interlayer charge transfer in the quantum Hall regime gives rise to novel phenomena with potential applications in cryogenic magnetometry.

ABSTRACT Railway transportation offers significant advantages for low-carbon mobility, provided that infrastructure integrity can be efficiently maintained with minimal service interruptions. This study presents a real-time rail defect detection system combining a three-phase excitation eddy current sensor array with a lightweight Transformer-based deep learning model. The sensor features a hexagonal array arrangement that achieves strong background signal cancellation and high lift-off tolerance, enabling reliable operation under dynamic conditions. Real-time conductivity mapping is performed through a rapid linear back-projection algorithm, and defect classification is accomplished using a Transformer model trained on COMSOL-simulated datasets. Experimental validation demonstrates a classification accuracy exceeding 95%, reliable identification of defects spaced as closely as 5 mm apart, and robustness across lift-off variations up to 10 mm. The system achieves an inference time of less than 0.02 seconds, supporting high-speed rail inspection at velocities up to 10.8 km/h without compromising detection performance. By facilitating continuous rail monitoring and reducing inspection-related energy consumption, the proposed system enhances the operational sustainability of railway infrastructure and contributes to broader low-carbon transportation initiatives. Future work will focus on expanding detection capabilities to subsurface and complex defects, further strengthening the system’s applicability for sustainable rail operation.

In this study, we successfully synthesized NbC nanofilms on the GaN surface, and a more uniform size and thinner thickness of NbC were optimized by further fabricating circular hole arrays on GaN epitaxial wafers using photolithography and etching techniques. This sensor exhibits an ultralow detection limit of 200 ppb for TMA gas at room temperature, a high response value (84.14%) to 200 ppm of TMA, and low resistance fluctuation for uniform NbC nanofilms. The excellent performance after combination of the two can be attributed to the synergistic effects of p-n heterojunctions and Schottky barriers. Furthermore, the algorithm innovatively adopts dual-feature extraction via KPCA combined with polynomial feature engineering to systematically investigate the relationship within sensor array data. By integrating machine learning algorithms with the sensor array, the system achieves the precise identification of target components in gas mixtures, reaching 98% accuracy. Ultimately, this study demonstrates the significant application potential of gas sensors in the next generation robotic electronic nose.

Here, we introduce Bi2SiO5 (BSO), an intrinsically UV-B selective oxide with a 4.17 eV bandgap (297 nm), and present a cost-effective, scalable conversion approach for its synthesis. A single-step reaction between Bi2O3 vapor and SiO2 in fused silica converts the glass directly into crystalline BSO. To enable scalable integration, SiO2 is first patterned on Al2O3; the same conversion then proceeds exclusively within those patterns, yielding pixel-defined BSO films with sharply defined edges. Devices fabricated on these films exhibit a UV-B photoresponse more than 360-fold stronger than their UV-A response, evidencing strong spectral selectivity. Owing to the clean selective-area conversion, arrays of devices show a relative standard deviation in responsivity of good 5.3%, underlining excellent uniformity and reproducibility. The synergy of intrinsic UV-B selectivity, high material quality, and patternable growth positions BSO as a compelling platform for next-generation UV-B photodetectors and integrated optical sensor arrays.

Colorimetric sensor array based on a bimetallic PtPd nanozyme for simultaneous discrimination of multiple antioxidants.

Self-healing and adhesive eutectogels with dual conductive networks for multimodal health monitoring and human-computer interaction.

Purpose Cable eccentricity not only causes electric field distortion and insulation layer aging, leading to a decline in cable performance, but also seriously threatens the safe operation of power systems. However, existing detection methods have problems such as insufficient accuracy or poor environmental adaptability. This study aims to present a cable eccentricity detection method based on an array eddy current sensor (ECS) for measuring the eccentricity of power cables. Design/methodology/approach Firstly, by designing an array model, the calculation formulas for the array ECS to detect cable eccentricity and eccentric angle were analyzed and derived. Secondly, a finite element simulation model of the sensor and multicore cable was established, and the coupling relationship between the sensor and the cable under scenarios with different numbers of array sensors was discussed, further verifying the feasibility of this method. Finally, static and dynamic measurement experiments on multicore cables were carried out using a cable eccentricity detection device. Findings The experimental results show that within the eccentricity range of 0–3.4 mm, the calculation error of eccentricity is within ±3% and the error of eccentric angle is within ±5%, which meets the accuracy requirements. Originality/value The designed array ECS can realize the measurement of cable eccentricity and offset angle, and has certain research significance.

Next-generation fire safety systems demand precise detection and motion recognition of flames. In-sensor computing, which integrates sensing, memory, and processing capabilities, has emerged as a key technology in flame detection. However, the implementation of hardware-level functional demonstrations based on artificial vision systems in the solar-blind ultraviolet (UV) band (200-280nm) is hindered by the weak detection capability. Here, we propose Ga2O3/In2Se3 heterojunctions for the ferroelectric (abbreviation: Fe) optoelectronic sensor (abbreviation: OES) array (5 × 5 pixels), which is capable of ultraweak UV light detection with an ultrahigh detectivity through ferroelectric regulation and features in configurable multimode functionality. The Fe-OES array can directly sense different flame motions and simulate the non-spiking gradient neurons of insect visual system. Moreover, the flame signal can be effectively amplified in combination with leaky integration-and-fire neuron hardware. Using this Fe-OES system and neuromorphic hardware, we successfully demonstrate three flame processing tasks: achieving efficient flame detection across all time periods with terminal and cloud-based alarms; flame motion recognition with a lightweight convolutional neural network achieving 96.47% accuracy; and flame light recognition with 90.51% accuracy by means of a photosensitive artificial neural system. This work provides effective tools and approaches for addressing a variety of complex flame detection tasks.

Intelligent food freshness monitoring systems are in urgent demand for food safety and quality assurance in modern cold chain transportation. Conventional monitoring techniques are rendered impractical within the hermetic confines of refrigeration units. Conversely, the utilization of electronic olfactory systems to discern distinct volatile molecules emitted by food stored during the deterioration process presents an effective monitoring remedy. In this study, an artificial olfactory chip featuring a nine-box-grid configured gas sensor array (9-Array) has been designed for smart cold chain management. Through its nanotubular design, the 9-Array chip exhibits remarkable sensitivity and discriminability to ammonia (NH3), ethanol (EtOH), trimethylamine (TMA), and hydrogen sulfide (H2S) emitted from decaying meat, vegetables, and fruits. Notably, this chip excels in accurately categorizing these gases and their combinations by leveraging distinct responses from individual sensor pixels, enabling the precise differentiation of overlapping gas signatures originating from diverse food sources. When seamlessly integrated into a food freshness monitoring system within a refrigerated environment, the 9-Array chip showcases reliable real-time monitoring capabilities and sustained performance over extended periods. It possesses the capability to recognize at least 7 types of foods and evaluate their freshness over 7 days. Moreover, our olfactory chip adeptly identifies and evaluates the freshness of varied food items within mixed food contexts, showing its potential as a future key component for advanced smart cold chain food management platforms that elevate food management efficiency and safety standards.

Nanozymes show great potential in environmental monitoring, yet their catalytic performance and recyclability remain challenging. This study developed a supported nanozyme system by pyrolyzing precursors to synthesize ultrafine CuO nanoclusters anchored on hollow sea-urchin-shaped SiO2 (CuO@SiO2). The material exhibits switchable peroxidase- and laccase-like activities, strong substrate affinity, and excellent recyclability, attributable to its unique crystal/amorphous hybrid structure. Leveraging this dual-enzyme activity, we constructed a dual-channel sensor array based on the "turn-on" laccase-like and "turn-off" peroxidase-like responses, enabling highly selective discrimination among dihydroxybenzene isomers at concentrations as low as 5 μM. The array successfully identified single isomers, mixed samples, and nine unknown analytes with 100% accuracy. This work provides new insights into supported nanozyme design and demonstrates a multidimensional signal extraction strategy to advance nanozyme sensing technology.

The conventional reactive cooling strategy, which relies on static thresholds, has become inadequate for managing dynamically changing heat loads, often resulting in energy inefficiency and increased risk of local hot spots. In this study, we develop a data center cooling optimization system that integrates distributed sensor arrays for predictive analysis. By deploying high-density temperature and humidity sensors both inside and outside server racks, a real-time, high-fidelity three-dimensional digital twin of the data center’s thermal environment is constructed. Time-series analysis combined with Long Short-Term Memory algorithms is employed to forecast temperature and humidity based on the extensive environmental data collected, achieving high predictive accuracy with a root mean square error of 0.25 and an R2 value of 0.985. Building on these predictions, a proactive cooling control strategy is formulated to dynamically adjust fan speeds and the opening degree of chilled-water valves in computer room air conditioning units, changing the cooling approach from passive to preemptive prevention of overheating. Compared with conventional proportional–integral–differential control, the developed system significantly reduces overall energy consumption and maintains all equipment within safe operating temperatures. Specifically, the framework has reduced the energy consumption of the cooling system by 37.5%, lowered the overall power usage effectiveness of the data center by 12% (1.48 to 1.30), and suppressed the cumulative hotspot duration (temperature 27 °C) by nearly 96% (from 48 to 2 h).

Honeycomb-like mesoporous Fe-N-C single-atom nanozyme with excellent POD-like activity: Preparation, properties, and application as a sensor array for discrimination of phenolic acids.

Globally, over 30% of patients with epilepsy (PWE) suffer from psychiatric comorbidities. The concurrent use of antiepileptic medications and psychotropic drugs in their treatment presents certain risks, such as potential adverse drug reactions or a reduction in therapeutic efficacy. However, most existing approaches are confined to single-drug analysis and lack the capability for the simultaneous detection and discrimination of multiple drugs. Hereon, a triple-channel array was fabricated using carbon dots (TS-CDs), with the assistance of Fe3+ and Cu2+, for simultaneously distinguishing six drugs, including two antiseizure medications and four psychotropic drugs. Notably, assisted by appropriate statistical analysis methods, the sensor array correctly identified all blind samples and achieved quantitative detection of all six drugs in human serum samples. This research establishes a promising approach for therapeutic drug monitoring in patients with comorbidities requiring polypharmacy.

Flexible wearable sensors have demonstrated tremendous potential in personalized health monitoring, but their signal acquisition accuracy critically depends on the dynamically changing contact quality between the sensor and skin. Existing flexible sensors predominantly employ passive adaptation strategies, which struggle to actively optimize contact states in complex physiological scenarios such as motion and sweating, leading to signal quality degradation or even monitoring failure. To address this challenge, this study proposes an active morphology control method for flexible sensor arrays based on photo-driven lattice metamaterials. We constructed a truncated octahedral lattice structure with both high flexibility and photothermal responsiveness using poly(N-isopropylacrylamide)-single-walled carbon nanotube (PNIPAM-SWNT) hydrogel, and integrated conductive polymer (PEDOT:PSS) as sensing electrodes. Through precise scanning of 780nm femtosecond laser, local controllable deformation of the sensor array can be achieved. Experimental results show that the lattice structure design reduces the relative density of the sensor array to 18.5%, with photo-driven deformation amplitude reaching 35.2% and response speed of 12.8 μm/s. Through various morphology control modes such as local protrusion, bending, and torsion, the contact impedance between sensor and skin was effectively reduced by 62.3%, and the signal-to-noise ratio (SNR) of electrocardiogram (ECG) signals improved by 18.7 dB during motion states. This study innovatively extends photo-driven soft material technology from the microrobotics field to wearable electronics, providing a new paradigm for developing next-generation intelligent wearable devices with active adaptation capabilities, embodying deep interdisciplinary integration of design, materials, electronics, and biomedicine.

In this study, a ThT@Q[8] host-guest complex was obtained via an interaction between cucurbit[8]uril (Q[8]) as the host and thioflavin T (ThT) as the guest. Then, ThT@Q[8]-Fe3+ and ThT@Q[8]-Hg2+ fluorescent probes were constructed based on metal ion stimuli-responsive supramolecular interactions. These probes exhibited distinct fluorescence responses toward five physiological phosphate anions (Pi, PPi, AMP, ADP, and ATP) due to electrostatic and coordinative interactions. A supramolecular fluorescence sensing array was thus developed using ThT@Q[8]-Fe3+ and ThT@Q[8]-Hg2+ as sensing units for the discrimination and detection of these phosphate anions. The array showed excellent anti-interference performance and successfully distinguished phosphate anions at low, medium, and high concentrations. A good linear relationship was observed between Factor 1 and anion concentration over the range from the limit of quantitation (LOQ) to 10 μM, with a limit of detection (LOD) as low as 0.13 μM. The array was also effective in discriminating the five phosphate anions in human serum. With advantages such as rich output signals, low cost, and simple preparation, this array holds considerable potential for detecting phosphate anions in biological systems. Furthermore, it enabled monitoring of ATP hydrolysis and evaluation of the energy charge in human serum.

Trinuclear copper complexes that emulate the active sites of multicopper oxidases (MCOs) are of broad biological interest. Here, a monoatomic-node strategy combined with solvent reduction was used to construct a rare trinuclear copper MOF featuring a planar, equilateral-triangular CuI 0.6CuII 2.4 core. The CuI 0.6CuII 2.4-MOF nanozyme shows enzymatic activity 37.2 times than that of laccase and 13.6 times than that of a CuII 3-MOF analogue, retains high chemical stability from pH 4 to 12, and lowers production cost by 16.2-fold relative to natural laccase. DFT calculations attribute the superior performance to the introduction of Cu+, which yields a more favorable electronic structure, reaction energy landscape, and intermediate binding than in CuII 3-MOF. Consequently, CuI 0.6CuII 2.4-MOF efficiently degrades phenolic pollutants and enables colorimetric discrimination and detection of 11 phenols as a colorimetric sensor array. It also affords sensitive detection of epinephrine and dopamine, with limits of detection of 6.5 and 10µm, respectively. Overall, this work demonstrates a laccase-inspired route to high-activity, low-cost MOFs with strong potential for environmental remediation and biosensing.

Reconstruction accuracy is a crucial indicator for assessing the combustion conditions in combustion diagnostics. Existing flame-temperature measurement methods require comparison and verification using simulations, infrared thermal imaging cameras, and thermocouples, which increase measurement complexity and reduce efficiency. This study introduces a method based on tunable diode laser absorption spectroscopy (TDLAS) that utilizes an array sensor to capture specific absorption and radiation spectral signals to measure the combustion field temperature. Active and passive spectral information at different temperatures was obtained through calibration experiments. The temperature field of the K+-doped premixed stable flame was imaged. The difference between the active and passive spectral temperature field images did not exceed 8%, and the maximum relative error compared to the thermocouple measurements did not exceed 6%, demonstrating good reconstruction accuracy. Simultaneous active and passive spectral temperature measurements provide a novel approach for developing high-precision flame measurements and reducing system complexity.

Accurate antioxidant detection remains crucial for biomedical and food applications, although conventional methods face persistent challenges including structural similarity interference, instrument dependency, and unreliable detection outcomes. To address these limitations, this study developed innovative nanozyme-based multisubstrate sensor array integrated with machine learning. A novel nanocomposite was constructed through immobilizing manganese- and cerium-codoped carbon dots (Mn,Ce-CDs) onto metal-organic frameworks MIL-53(Fe). The Mn,Ce-CDs significantly enhanced electron transfer efficiency and remarkably improved superoxide anion (O2•-)-mediated oxidase-like activity of the composite. This nanozyme effectively catalyzed four chromogenic substrates, 3,3',5,5'-tetramethylbenzidine (TMB), o-phenylenediamine (OPD), 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS), and 4-aminoantipyrine (4-AP), to generate distinct multicolor fingerprint patterns, enabling multidimensional signal outputs without requiring exogenous metal ions or multiple probes. The research incorporated seven machine learning algorithms, including linear discriminant analysis (LDA), hierarchical clustering analysis (HCA), artificial neural network (ANN), K-nearest neighbors (KNN), support vector machine (SVM), decision tree (DT), and naive bayes (NB), leveraging their discriminative criterion variations to establish complementary effects that successfully differentiated eight structurally analogous antioxidants (10 nM to 25 μM), with a minimum identification concentration of 10 nM. To rectify misclassification arising from DT and NB algorithms' sensitivity to high-dimensional features, innovative LDA-DT and LDA-NB tandem algorithms were developed. Through employing dimensionality reduction techniques, the classification accuracy of the two algorithms was significantly improved from 86.7% and 93.1% to 100%, respectively. The developed sensor array successfully enabled quantitative analysis of the eight antioxidants and achieved accurate identification of multiple antioxidants in complex matrices including serum, urine, cell lysates, bacterial cultures, and various food samples.

Engineering the microenvironment of Cu-MOF nanozyme via modulating ligand hydrophobicity for array-based profiling of phenolic acids in natural products.