Sensors

19 pages, 1680 KB

Open AccessArticle

An Improved DQN Framework with Dual Residual Horizontal Feature Pyramid for Autonomous Fault Diagnosis in Strong-Noise Scenarios

by Sha Li, Tong Wang, Xin Xu, Weiting Gan, Kun Chen, Xinyan Fan and Xueming Xu

Sensors 2025, 25(24), 7639; https://doi.org/10.3390/s25247639 - 16 Dec 2025

Abstract

Fault diagnosis methods based on deep learning have made certain progress in recent years. However, in actual industrial scenarios, there are severe strong background noise and limited computing resources, which poses challenges to the practical application of fault diagnosis models. In response to [...] Read more.

Fault diagnosis methods based on deep learning have made certain progress in recent years. However, in actual industrial scenarios, there are severe strong background noise and limited computing resources, which poses challenges to the practical application of fault diagnosis models. In response to the above issues, this paper proposes a novel noise-resistant and lightweight fault diagnosis framework with nonlinear timestep degenerative greedy strategy (NTDGS) and dual residual horizontal feature pyramid (DRHFPN) for fault diagnosis in strong noise scenarios. This method takes advantage of the strong fitting ability of deep learning methods to model the agent in reinforcement learning by the ways of parameterization, fully leveraging the advantages of both deep learning and reinforcement learning methods. NTDGS is further developed to adaptively adjust the action sampling strategy of the agent at different training stages, improving the convergence speed of the network. To enhance the noise resistance of the network, DRHFPN is constructed, which can filter out interference noise at the feature map level by fusing local feature details and global semantic information. Furthermore, the feature map weighting attention mechanism (FMWAM) is designed to enhance the weak feature extraction ability of the network through adaptive weighting of the feature maps. Finally, the performance of the proposed method is evaluated in different datasets and strong noise environments. Experiments show that in various fault diagnosis scenarios, the proposed method has better noise resistance, higher fault diagnosis accuracy, and fewer parameters compared to other methods. Full article

(This article belongs to the Special Issue Smart Sensors for Machine Condition Monitoring and Fault Diagnosis)

31 pages, 5270 KB

Open AccessArticle

Multi-Serial Adaptive Bus Interface with Integrated Monitoring and Plug-And-Play Connectivity

by Marcel Tresanchez and Tomàs Pallejà

Sensors 2025, 25(24), 7638; https://doi.org/10.3390/s25247638 - 16 Dec 2025

Abstract

This work presents a complete multi-serial adaptive bus interface system compatible with the most widely used industrial serial communications standards: RS-232, RS-485, RS-422, and CAN. The proposed system automatically detects the connected serial interface type through analog line sensors and dynamically redirects the [...] Read more.

This work presents a complete multi-serial adaptive bus interface system compatible with the most widely used industrial serial communications standards: RS-232, RS-485, RS-422, and CAN. The proposed system automatically detects the connected serial interface type through analog line sensors and dynamically redirects the bus to the appropriate transceiver using a logical multiplexer. This approach aims to simplify the configuration of heterogeneous serial devices in complex and modular integration scenarios, such as body builders in industrial or vehicular systems. The hardware is designed as a scalable PCIe card-based device, allowing multiple adaptive bus interfaces to be integrated within a rack-based modular architecture. In addition, a single 5-pin plug-and-play connector is proposed by unifying the different bus signals of the transceivers, thereby simplifying cabling and deployment. Complementary implemented capabilities include baud rate auto-detection and supervision, as well as automatic direction-control functionality for RS-485 communication. Experimental validation demonstrated that the proposed system successfully detected and redirected all supported interfaces, achieving reliable connection and disconnection within an average time of 2.5 s. Furthermore, the integrated baud rate auto-detection algorithm accurately identified transmission speeds up to 1 Mbps in under 80 ms, while the automatic direction-control capability operated reliably at speeds up to 576,000 bps. Full article

(This article belongs to the Special Issue Joint Communication and Sensing in Vehicular Networks)

► Show Figures

Figure 1

20 pages, 3287 KB

Open AccessArticle

Dual-Branch Superpixel and Class-Center Attention Network for Efficient Semantic Segmentation

by Yunting Zhang, Hongbin Yu, Haonan Wang, Mengru Zhou, Tao Zhang and Yeh-Cheng Chen

Sensors 2025, 25(24), 7637; https://doi.org/10.3390/s25247637 - 16 Dec 2025

Abstract

With the advancement of deep learning, image semantic segmentation has achieved remarkable progress. However, the complexity and real-time requirements of practical applications pose greater challenges for segmentation algorithms. To address these, we propose a dual-branch network guided by attention mechanisms that tackles common [...] Read more.

With the advancement of deep learning, image semantic segmentation has achieved remarkable progress. However, the complexity and real-time requirements of practical applications pose greater challenges for segmentation algorithms. To address these, we propose a dual-branch network guided by attention mechanisms that tackles common limitations in existing methods, such as coarse edge segmentation, insufficient contextual understanding, and high computational overhead. Specifically, we introduce a superpixel sampling weighting module that models pixel dependencies based on different regional affiliations, thereby enhancing the network’s sensitivity to object boundaries while preserving local features. Furthermore, a class-center attention module is designed to extract class-centered features and facilitate category-aware modeling. This module reduces the computational overhead and redundancy of traditional self-attention mechanisms, thereby improving the network’s global feature representation. Additionally, learnable parameters are employed to adaptively fuse features from both branches, enabling the network to better focus on critical information. We validate our method on three benchmark datasets (PASCAL VOC 2012, Cityscapes, and ADE20K) by comparing it with mainstream models including FCN, DeepLabV3+, and DANet, with evaluation metrics of mIoU and PA. Our method delivers superior segmentation performance in these experiments. These results underscore the effectiveness of the proposed algorithm in balancing segmentation accuracy and model efficiency. Full article

(This article belongs to the Section Sensing and Imaging)

► Show Figures

Figure 1

74 pages, 2205 KB

Open AccessSystematic Review

Artificial Intelligence of Things for Next-Generation Predictive Maintenance

by Taimia Bitam, Aya Yahiaoui, Djallel Eddine Boubiche, Rafael Martínez-Peláez, Homero Toral-Cruz and Pablo Velarde-Alvarado

Sensors 2025, 25(24), 7636; https://doi.org/10.3390/s25247636 - 16 Dec 2025

Abstract

Industry 5.0 introduces a shift toward human-centric, sustainable, and resilient industrial ecosystems, emphasizing intelligent automation, collaboration, and adaptive operations. Predictive Maintenance (PdM) plays a critical role in this transition, addressing the limitations of traditional maintenance approaches in increasingly complex and data-driven environments. The [...] Read more.

Industry 5.0 introduces a shift toward human-centric, sustainable, and resilient industrial ecosystems, emphasizing intelligent automation, collaboration, and adaptive operations. Predictive Maintenance (PdM) plays a critical role in this transition, addressing the limitations of traditional maintenance approaches in increasingly complex and data-driven environments. The convergence of Artificial Intelligence and the Industrial Internet of Things, referred to as the Artificial Intelligence of Things (AIoT), enables real-time sensing, learning, and decision-making for advanced fault detection, Remaining Useful Life estimation, and prescriptive maintenance actions. This study provides a systematic and structured review of AIoT-enabled PdM aligned with Industry 5.0 objectives. It presents a unified taxonomy integrating AI models, Industrial Internet of Things (IIoT) infrastructures, and AIoT architectures; reviews AI-driven techniques, sector-specific implementations in manufacturing, transportation, and energy; and analyzes emerging paradigms such as Edge–Cloud collaboration, federated learning, self-supervised learning, and digital twins for autonomous and privacy-preserving maintenance. Furthermore, this paper synthesizes strengths, limitations, and cross-industry challenges, and outlines future research directions centered on explainability, data quality and heterogeneity, resource-constrained intelligence, cybersecurity, and human–AI collaboration. By bridging technological advancements with Industry 5.0 principles, this review contributes a comprehensive foundation for the development of scalable, trustworthy, and next-generation AIoT-based predictive maintenance systems. Full article

(This article belongs to the Section Internet of Things)

18 pages, 907 KB

Open AccessArticle

Towards Generalized Bioimpedance Models for Bladder Monitoring: The Role of Waist Circumference and Fat Thickness

by H. Trask Crane, John A. Berkebile, Samer Mabrouk, Nicholas Riccardelli and Omer T. Inan

Sensors 2025, 25(24), 7635; https://doi.org/10.3390/s25247635 - 16 Dec 2025

Abstract

Continuous bladder volume monitoring in a wearable format can improve outcomes for patients with bladder dysfunction, heart failure, and other conditions requiring precise fluid management. Bioimpedance-based methods offer a promising, noninvasive solution; however, the influence of patient-specific anatomy, particularly waist circumference and subcutaneous [...] Read more.

Continuous bladder volume monitoring in a wearable format can improve outcomes for patients with bladder dysfunction, heart failure, and other conditions requiring precise fluid management. Bioimpedance-based methods offer a promising, noninvasive solution; however, the influence of patient-specific anatomy, particularly waist circumference and subcutaneous fat thickness, remains poorly characterized. In this study, we use in silico finite element modeling to quantify how these anatomical factors affect two key bioimpedance metrics: voltage change (

Δ V

) and voltage change ratio (VCR). Comprehensive simulations were performed across 15 virtual anatomies, generating a reference dataset for guiding future analog front-end and algorithm designs. We further compared generalized volume estimation models against conventional patient-specific void regression approaches. With appropriate input scaling, the generalized models achieved performance within 10% of patient-specific calibrations and, in some cases, surpassed them. Certain configurations reduced mean average error (MAE) by more than 20% relative to individualized models, potentially enabling a streamlined setup without the need for laborious ground-truth acquisition such as voided volume collection. These results demonstrate that incorporating simple anatomical scaling can yield robust, generalizable bladder volume estimation models suitable for wearable systems across diverse patient populations. Full article

(This article belongs to the Special Issue Advanced Sensors for Human Health Management)

25 pages, 2228 KB

Open AccessArticle

EEG Sensor-Based Computational Model for Personality and Neurocognitive Health Analysis Under Social Stress

by Majid Riaz, Pedro Guerra and Raffaele Gravina

Sensors 2025, 25(24), 7634; https://doi.org/10.3390/s25247634 - 16 Dec 2025

Abstract

This paper introduces an innovative EEG sensor-based computational framework that establishes a pioneering nexus between personality trait quantification and neural dynamics, leveraging biosignal processing of brainwave activity to elucidate their intrinsic influence on cognitive health and oscillatory brain rhythms. By employing electroencephalography (EEG) [...] Read more.

This paper introduces an innovative EEG sensor-based computational framework that establishes a pioneering nexus between personality trait quantification and neural dynamics, leveraging biosignal processing of brainwave activity to elucidate their intrinsic influence on cognitive health and oscillatory brain rhythms. By employing electroencephalography (EEG) recordings from 21 participants undergoing the Trier Social Stress Test (TSST), we propose a machine learning (ML)-driven methodology to decode the Big Five personality traits—Extraversion (Ex), Agreeableness (A), Neuroticism (N), Conscientiousness (C), and Openness (O)—using classification algorithms such as support vector machine (SVM) and multilayer perceptron (MLP) applied to 64-electrode EEG sensor data. A novel multiphase neurocognitive analysis across the TSST stages (baseline, mental arithmetic, job interview, and recovery) systematically evaluates the bidirectional relationship between personality traits and stress-induced neural responses. The proposed framework reveals significant negative correlations between frontal–temporal theta–beta ratio (TBR) and self-reported Extraversion, Conscientiousness, and Openness, indicating faster stress recovery and higher cognitive resilience in individuals with elevated trait scores. The binary classification model achieves high accuracy (88.1% Ex, 94.7% A, 84.2% N, 81.5% C, and 93.4% O), surpassing the current benchmarks in personality neuroscience. These findings empirically validate the close alignment between personality constructs and neural oscillatory patterns, highlighting the potential of EEG-based sensing and machine-learning analytics for personalized mental-health monitoring and human-centric AI systems attuned to individual neurocognitive profiles. Full article

(This article belongs to the Section Internet of Things)

► Show Figures

Figure 1

26 pages, 1566 KB

Open AccessArticle

Improved Moth-Inspired Algorithm Based on Fuzzy Controller

by Zhoujing Lv, Dongxu Liu, Yu Wu and Huichao Zhu

Sensors 2025, 25(24), 7633; https://doi.org/10.3390/s25247633 - 16 Dec 2025

Abstract

In recent years, the demand for mobile robots to perform odor source localization tasks in dangerous environments has been increasing, and this technology has gradually become a focus of research. However, existing bio-inspired algorithms still have many limitations in real applications. Therefore, we [...] Read more.

In recent years, the demand for mobile robots to perform odor source localization tasks in dangerous environments has been increasing, and this technology has gradually become a focus of research. However, existing bio-inspired algorithms still have many limitations in real applications. Therefore, we attempt to design a moth-inspired algorithm integrated with a fuzzy control mechanism to enhance the robot’s ability to track odor sources in environments with dense obstacles. The goal is to improve the accuracy and efficiency of localization. Through multiple sets of simulation and real environment experiments, we observed that this algorithm achieved significant improvements in multiple indicators, including task success rate, search efficiency, and path planning quality. Compared with the traditional moth algorithm, its stability and adaptability in complex scenarios are also outstanding. Full article

(This article belongs to the Special Issue Advancements and Applications of Biomimetic Sensors Technologies)

31 pages, 1771 KB

Open AccessArticle

Forecasting Energy Demand in Quicklime Manufacturing: A Data-Driven Approach

by Jersson X. Leon-Medina, John Erick Fonseca Gonzalez, Nataly Yohana Callejas Rodriguez, Mario Eduardo González Niño, Saúl Andrés Hernández Moreno, Wilman Alonso Pineda-Munoz, Claudia Patricia Siachoque Celys, Bernardo Umbarila Suarez and Francesc Pozo

Sensors 2025, 25(24), 7632; https://doi.org/10.3390/s25247632 - 16 Dec 2025

Abstract

This study presents a deep learning-based framework for forecasting energy demand in a quicklime production company, aiming to enhance operational efficiency and enable data-driven decision-making for industrial scalability. Using one year of real electricity consumption data, the methodology integrates temporal and operational variables—such [...] Read more.

This study presents a deep learning-based framework for forecasting energy demand in a quicklime production company, aiming to enhance operational efficiency and enable data-driven decision-making for industrial scalability. Using one year of real electricity consumption data, the methodology integrates temporal and operational variables—such as load profile, active power, shift indicators, and production-related proxies—to capture the dynamics of energy usage throughout the manufacturing process. Several neural network architectures, including Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Conv1D models, were trained and compared to predict short-term power demand with 10-min resolution. Among these, the GRU model achieved the highest predictive accuracy, with a best performance of RMSE = 2.18 kW, MAE = 0.49 kW, and SMAPE = 3.64% on the test set. The resulting forecasts support cost-efficient scheduling under time-of-use tariffs and provide valuable insights for infrastructure planning, capacity management, and sustainability optimization in energy-intensive industries. Full article

(This article belongs to the Section Fault Diagnosis & Sensors)

► Show Figures

Figure 1

20 pages, 1736 KB

Open AccessArticle

Experimental Investigation of One-Way Lamb and SH Mixing Method in Composite Laminates

by Siyang Xie, Youxuan Zhao and Yuzi Liu

Sensors 2025, 25(24), 7631; https://doi.org/10.3390/s25247631 - 16 Dec 2025

Abstract

This paper experimentally investigates the resonant behavior of the one-way Lamb and SH (shear horizontal) mixing method in composite laminates with inherent quadratic nonlinearity, delamination damage and impact damage. When the fundamental S₀-mode Lamb waves and SH₀ waves mix in [...] Read more.

This paper experimentally investigates the resonant behavior of the one-way Lamb and SH (shear horizontal) mixing method in composite laminates with inherent quadratic nonlinearity, delamination damage and impact damage. When the fundamental S₀-mode Lamb waves and SH₀ waves mix in the damage regions of composite laminates, experimental results demonstrate the generation of the resonant SH₀ waves with the resonance condition. Meanwhile, the damage localization method in composite laminates is experimentally verified by the time-domain signal of resonant waves. Furthermore, it is found that the one-way Lamb and SH mixing method is sensitive to inherent quadratic nonlinearity and impact damage but insensitive to delamination damage. Full article

(This article belongs to the Section Fault Diagnosis & Sensors)

22 pages, 12312 KB

Open AccessArticle

ES-YOLO: Multi-Scale Port Ship Detection Combined with Attention Mechanism in Complex Scenes

by Lixiang Cao, Jia Xi, Zixuan Xie, Teng Feng and Xiaomin Tian

Sensors 2025, 25(24), 7630; https://doi.org/10.3390/s25247630 - 16 Dec 2025

Abstract

With the rapid development of remote sensing technology and deep learning, the port ship detection based on a single-stage algorithm has achieved remarkable results in optical imagery. However, most of the existing methods are designed and verified in specific scenes, such as fixed [...] Read more.

With the rapid development of remote sensing technology and deep learning, the port ship detection based on a single-stage algorithm has achieved remarkable results in optical imagery. However, most of the existing methods are designed and verified in specific scenes, such as fixed viewing angle, uniform background, or open sea, which makes it difficult to deal with the problem of ship detection in complex environments, such as cloud occlusion, wave fluctuation, complex buildings in the harbor, and multi-ship aggregation. To this end, ES-YOLO framework is proposed to solve the limitations of ship detection. A novel edge perception channel, Spatial Attention Mechanism (EACSA), is proposed to enhance the extraction of edge information and improve the ability to capture feature details. A lightweight spatial–channel decoupled down-sampling module (LSCD) is designed to replace the down-sampling structure of the original network and reduce the complexity of the down-sampling stage. A new hierarchical scale structure is designed to balance the detection effect of different scale differences. In this paper, a remote sensing ship dataset, TJShip, is constructed based on Gaofen-2 images, which covers multi-scale targets from small fishing boats to large cargo ships. The TJShip dataset was adopted as the data source, and the ES-YOLO model was employed to conduct ablation and comparison experiments. The results show that the introduction of EACSA attention mechanism, LSCD, and multi-scale structure improves the mAP of ship detection by 0.83%, 0.54%, and 1.06%, respectively, compared with the baseline model, also performing well in precision, recall and F1. Compared with Faster R-CNN, RetinaNet, YOLOv5, YOLOv7, and YOLOv8 methods, the results show that the ES-YOLO model improves the mAP by 46.87%, 8.14%, 1.85%, 1.75%, and 0.86%, respectively, under the same experimental conditions, which provides research ideas for ship detection. Full article

(This article belongs to the Section Remote Sensors)

► Show Figures

Figure 1

31 pages, 11486 KB

Open AccessArticle

Towards Heart Rate Estimation in Complex Multi-Target Scenarios: A High-Precision FMCW Radar Scheme Integrating HDBS and VLW

by Xuefei Dong, Yunxue Liu, Jinwei Wang, Shie Wu, Chengyou Wang and Shiqing Tang

Sensors 2025, 25(24), 7629; https://doi.org/10.3390/s25247629 - 16 Dec 2025

Abstract

Non-contact heart rate estimation technology based on frequency-modulated continuous wave (FMCW) radar has garnered extensive attention in single-target scenarios, yet it remains underexplored in multi-target environments. Accurate discrimination of multiple targets and precise estimation of their heart rates constitute key challenges in the [...] Read more.

Non-contact heart rate estimation technology based on frequency-modulated continuous wave (FMCW) radar has garnered extensive attention in single-target scenarios, yet it remains underexplored in multi-target environments. Accurate discrimination of multiple targets and precise estimation of their heart rates constitute key challenges in the multi-target domain. To address these issues, we propose a novel scheme for multi-target heart rate estimation. First, a high-precision distance-bin selection (HDBS) method is proposed for target localization in the range domain. Next, multiple-input multiple-output (MIMO) array processing is combined with the Root-multiple signal classification (Root-MUSIC) algorithm for angular domain estimation, enabling accurate discrimination of multiple targets. Subsequently, we propose an efficient method for interference suppression and vital sign extraction that cascades variational mode decomposition (VMD), local mean decomposition (LMD), and wavelet thresholding (WT) termed as VLW, which enables high-quality heartbeat signal extraction. Finally, to achieve high-precision and super-resolution heart rate estimation with low computational burden, an improved fast iterative interpolated beamforming (FIIB) algorithm is proposed. Specifically, by leveraging the conjugate symmetry of real-valued signals, the improved FIIB algorithm reduces the execution time by approximately 60% compared to the standard version. In addition, the proposed scheme provides sufficient signal-to-noise ratio (SNR) gain through low-complexity accumulation in both distance and angle estimation. Six experimental scenarios are designed, incorporating densely arranged targets and front-back occlusion, and extensive experiments are conducted. Results show this scheme effectively discriminates multiple targets in all tested scenarios with a mean absolute error (MAE) below 2.6 beats per minute (bpm), demonstrating its viability as a robust multi-target heart rate estimation scheme in various engineering fields. Full article

(This article belongs to the Topic Advanced Array Antenna Design and Signal Processing Techniques)

31 pages, 10197 KB

Open AccessArticle

A Wi-Fi/PDR Fusion Localization Method Based on Genetic Algorithm Global Optimization

by Linpeng Zhang, Ji Ma, Yanhua Liu, Lian Duan, Yunfei Liang and Yanhe Lu

Sensors 2025, 25(24), 7628; https://doi.org/10.3390/s25247628 - 16 Dec 2025

Abstract

In indoor environments, fusion localization methods that combine Wi-Fi fingerprinting and Pedestrian Dead Reckoning (PDR) are constrained by the high sensitivity of traditional filters, such as the Extended Kalman Filter (EKF), to initial states and by their susceptibility to nonlinear drift. This study [...] Read more.

In indoor environments, fusion localization methods that combine Wi-Fi fingerprinting and Pedestrian Dead Reckoning (PDR) are constrained by the high sensitivity of traditional filters, such as the Extended Kalman Filter (EKF), to initial states and by their susceptibility to nonlinear drift. This study presents a Wi-Fi/PDR fusion localization approach based on global geometric alignment optimized via a Genetic Algorithm (GA). The proposed method models the PDR trajectory as an integrated geometric entity and performs a global search for the optimal two-dimensional similarity transformation that aligns it with discrete Wi-Fi observations, thereby eliminating dependence on precise initial conditions and mitigating multipath noise. Experiments conducted in a real office environment (14 × 9 m, eight dual-band APs) with a double-L trajectory demonstrate that the proposed GA fusion achieves the lowest mean error of 0.878 m (compared to 2.890 m, 1.277 m, and 1.193 m for Wi-Fi, PDR, and EKF fusion, respectively) and an RMSE of 0.978 m. It also attains the best trajectory fidelity (DTW = 0.390 m, improving by 71.0%, 14.7%, and 27.8%) and the smallest maximum deviation (Hausdorff = 1.904 m, 52.4% lower than Wi-Fi). The cumulative error distribution shows that 90% of GA fusion errors are within 1.5 m, outperforming EKF and PDR. Additional experiments that compare the proposed GA optimizer with Levenberg–Marquardt (LM), particle swarm optimization (PSO), and Procrustes alignment, as well as tests with 30% artificial Wi-Fi outliers, further confirm the robustness of the Huber-based cost and the effectiveness of the global optimization framework. These results indicate that the proposed GA-based fusion method achieves high robustness and accuracy in the tested office-scale scenario and demonstrate its potential as a practical multi-sensor fusion approach for indoor localization. Full article

(This article belongs to the Special Issue Smart Sensor Systems for Positioning and Navigation)

► Show Figures

Figure 1

20 pages, 2313 KB

Open AccessArticle

A Cybersecurity NER Method Based on Hard and Easy Labeled Training Data Discrimination

by Lin Ye, Yue Wu, Hongli Zhang and Mengmeng Ge

Sensors 2025, 25(24), 7627; https://doi.org/10.3390/s25247627 - 16 Dec 2025

Abstract

Although general-domain Named Entity Recognition (NER) has achieved substantial progress in the past decade, its application to cybersecurity NER is hindered by the lack of publicly available annotated datasets, primarily because of the sensitive and privacy-related nature of security data. Prior research has [...] Read more.

Although general-domain Named Entity Recognition (NER) has achieved substantial progress in the past decade, its application to cybersecurity NER is hindered by the lack of publicly available annotated datasets, primarily because of the sensitive and privacy-related nature of security data. Prior research has largely sought to improve performance by expanding annotation volumes, while overlooking the intrinsic characteristics of training data. In this study, we propose a cybersecurity Named Entity Recognition (NER) method based on hard and easy labeled training data discrimination. Firstly, a hybrid strategy that integrates a deep learning (DL)-based discriminator and a rule-based discriminator is employed to partition the original dataset into hard and easy samples. Secondly, the proportion of hard and easy data in the training set is adjusted to determine the optimal balance. Finally, a data augmentation algorithm is applied to the partitioned dataset to further improve model performance. The results demonstrate that, under a fixed total training data size, the ratio of hard to easy samples has a significant impact on NER performance, with the optimal strategy achieved at a 1:1 proportion. Moreover, the proposed method further improves the overall performance of cybersecurity NER. Full article

(This article belongs to the Special Issue Artificial Intelligence and Sensor Systems: From Attack Detection to Cyber Security)

► Show Figures

Figure 1

17 pages, 6578 KB

Open AccessArticle

Analysis of Wellbore Wall Deformation in Deep Vertical Wells Based on Fiber Bragg Grating Sensing Technology

by Wenchang Huang, Haibing Cai, Longfei Yang and Zixiang Li

Sensors 2025, 25(24), 7626; https://doi.org/10.3390/s25247626 - 16 Dec 2025

Abstract

Accurate deformation monitoring is essential for ensuring the stability of deep vertical shafts. In this study, a temperature-compensated fiber Bragg grating (FBG) sensing system was deployed in the 882 m deep Guotun Coal Mine shaft to measure circumferential and vertical strains at six [...] Read more.

Accurate deformation monitoring is essential for ensuring the stability of deep vertical shafts. In this study, a temperature-compensated fiber Bragg grating (FBG) sensing system was deployed in the 882 m deep Guotun Coal Mine shaft to measure circumferential and vertical strains at six depths. A site-specific mechanical model integrating stratigraphy, dual-layer concrete lining, and the influence radius was developed to analyze shaft wall stresses. The monitoring results reveal pronounced spatial anisotropy, with circumferential compressive and tensile strains at deeper levels nearly twice those at shallow levels. Strain variation also increases over time, reflecting the combined effects of groundwater fluctuations and overburden consolidation. The stresses inferred from measured strains agree well with the analytical solution in both magnitude and depth-dependent trend, with deviations remaining within a reasonable engineering margin. All stresses are below the strength limits of the C70/C50 concrete lining, confirming that the shaft is in a safe stress state. The proposed monitoring–analysis framework provides a reliable basis for evaluating shaft wall behavior under complex hydrogeological conditions. Full article

(This article belongs to the Section Optical Sensors)

► Show Figures

Figure 1

22 pages, 919 KB

Open AccessArticle

GeoCross: A Privacy-Preserving and Fine-Grained Authorization Scheme for Cross-Chain Geological Data Sharing

by Licheng Lin, Bin Feng and Pujie Jing

Sensors 2025, 25(24), 7625; https://doi.org/10.3390/s25247625 - 16 Dec 2025

Abstract

With the rapid development of geological blockchains and Internet of Things-based data acquisition technologies, massive amounts of heterogeneous data are constantly emerging. However, this data is stored in a distributed manner across different organizational or business blockchains. Data sharing among multiple geological blockchains [...] Read more.

With the rapid development of geological blockchains and Internet of Things-based data acquisition technologies, massive amounts of heterogeneous data are constantly emerging. However, this data is stored in a distributed manner across different organizational or business blockchains. Data sharing among multiple geological blockchains faces numerous challenges, either exposing sensitive data during verification or lacking effective authorization mechanisms. Therefore, how to achieve fine-grained access control and privacy protection across multiple blockchains has become a critical issue that must be addressed in geological data sharing. In this paper, we propose GeoCross, a cross-chain geological data sharing framework that enables fine-grained authorization management and privacy protection. First, GeoCross provides a hierarchical hybrid encryption mechanism that uses symmetric encryption for geological data protection and ciphertext-policy attribute-based encryption to enable flexible cross-chain access policies. Second, we integrate a Groth16-based zero-knowledge proof mechanism, which allows a chain to verify the existence, integrity, and accessibility of off-chain data without revealing the content. Furthermore, we introduce a Reputation-based Non-interactive Relay node Selection protocol (RNRS), which enhances the trustworthiness and fairness of cross-chain routing. Finally, we implement GeoCross in a multi-chain Hyperledger Fabric environment and evaluate its performance under real-world workloads. Results show that Groth16 verification requires only three bilinear pairings, achieving a throughput of up to 390 tps on a single chain and 1550 tps in a concurrent multi-chain environment. Even with 50% malicious nodes, the RNRS protocol still maintains a success rate of over 91%. These results demonstrate that GeoCross provides an efficient and practical solution for secure and privacy-preserving cross-chain geological data sharing. Full article

(This article belongs to the Special Issue Blockchain-Based Solutions to Secure IoT)

► Show Figures

Figure 1

21 pages, 16491 KB

Open AccessArticle

Glue Strips Measurement and Breakage Detection Based on YOLOv11 and Pixel Geometric Analysis

by Yukai Lu, Xihang Li, Jingran Kang, Shusheng Xiong and Shaopeng Zhu

Sensors 2025, 25(24), 7624; https://doi.org/10.3390/s25247624 - 16 Dec 2025

Abstract

With the rapid development of the new energy vehicle industry, the quality control of battery pack glue application processes has become a critical factor in ensuring the sealing, insulation, and structural stability of the battery. However, existing detection methods face numerous challenges in [...] Read more.

With the rapid development of the new energy vehicle industry, the quality control of battery pack glue application processes has become a critical factor in ensuring the sealing, insulation, and structural stability of the battery. However, existing detection methods face numerous challenges in complex industrial environments, such as metal reflections, interference from heating film grids, inconsistent orientations of glue strips, and the difficulty of accurately segmenting elongated targets, leading to insufficient precision and robustness in glue dimension measurement and glue break detection. To address these challenges, this paper proposes a battery pack glue application detection method that integrates the YOLOv11 deep learning model with pixel-level geometric analysis. The method first uses YOLOv11 to precisely extract the glue region and identify and block the heating film interference area. Glue strips orientation correction and image normalization are performed through adaptive binarization and Hough transformation. Next, high-precision pixel-level measurement of glue strip width and length is achieved by combining connected component analysis and multi-line statistical strategies. Finally, glue break and wire drawing defects are reliably detected based on image slicing and pixel ratio analysis. Experimental results show that the average measurement errors in glue strip width and length are only 1.5% and 2.3%, respectively, with a 100% accuracy rate in glue break detection, significantly outperforming traditional vision methods and mainstream instance segmentation models. Ablation experiments further validate the effectiveness and synergy of the modules. This study provides a high-precision and robust automated detection solution for glue application processes in complex industrial scenarios, with significant engineering application value. Full article

(This article belongs to the Section Sensing and Imaging)

► Show Figures

Figure 1

15 pages, 46278 KB

Open AccessArticle

Assessment of KN95 Mask Filtering Degradation and Breathing Detection: A Pilot Study

by Julie Payette, Alexandre Perrotton, Paul Fourmont, Fabrice Vaussenat, Jaime A. Benavides, Luis Felipe Gerlein and Sylvain G. Cloutier

Sensors 2025, 25(24), 7623; https://doi.org/10.3390/s25247623 - 16 Dec 2025

Abstract

This study aims to monitor mask performance in operando using all-printed humidity sensor arrays based on BiFeO₃/BiOCl heterostructures. Two screen-printed 19-sensor arrays are fixed directly atop the mask, in order to analyze moisture levels in exhaled breath and extract performance indicators. [...] Read more.

This study aims to monitor mask performance in operando using all-printed humidity sensor arrays based on BiFeO₃/BiOCl heterostructures. Two screen-printed 19-sensor arrays are fixed directly atop the mask, in order to analyze moisture levels in exhaled breath and extract performance indicators. This approach allows for an examination of the humidity saturation and absorption over time during operation. Accumulation of moisture within the mask can affect its performance, and factors like breath humidity, mask material, and ambient conditions influence this. Results show that the measured data follows an exponential decay, achieving correlation factors of over 0.9 for all tests. We also detect breathing differences through feature extraction, investigating the respiration rates and signal amplitudes for both normal and deep breathing. Furthermore, we animated the airflow in the mask in both 2D and 3D, allowing for the eventual detection of leaks for ill-fitting masks. This study introduces an innovative approach for the assessment of mask fit and longevity, contributing to improving mask efficacy and public health outcomes. Full article

(This article belongs to the Special Issue Sensors for Breathing Monitoring—2nd Edition)

► Show Figures

Figure 1

25 pages, 3766 KB

Open AccessArticle

WiFi RSS and RTT Indoor Positioning with Graph Temporal Convolution Network

by Lila Rana and Aayush Dulal

Sensors 2025, 25(24), 7622; https://doi.org/10.3390/s25247622 - 16 Dec 2025

Abstract

Indoor positioning using commodity WiFi has gained significant attention; however, achieving sub-meter accuracy across diverse layouts remains challenging due to multipath fading and Non-Line-Of-Sight (NLOS) effects. In this work, we propose a hybrid Graph–Temporal Convolutional Network (GTCN) model that incorporates Access Point (AP) [...] Read more.

Indoor positioning using commodity WiFi has gained significant attention; however, achieving sub-meter accuracy across diverse layouts remains challenging due to multipath fading and Non-Line-Of-Sight (NLOS) effects. In this work, we propose a hybrid Graph–Temporal Convolutional Network (GTCN) model that incorporates Access Point (AP) geometry through graph convolutions while capturing temporal signal dynamics via dilated temporal convolutional networks. The proposed model adaptively learns per-AP importance using a lightweight gating mechanism and jointly exploits WiFi Received Signal Strength (RSS) and Round-Trip Time (RTT) features for enhanced robustness. The model is evaluated across four experimental areas such as lecture theatre, office, corridor, and building floor covering areas from 15 m × 14.5 m to 92 m × 15 m. We further analyze the sensitivity of the model to AP density under both LOS and NLOS conditions, demonstrating that positioning accuracy systematically improves with denser AP deployment, especially in large-scale mixed environments. Despite its high accuracy, the proposed GTCN remains computationally lightweight, requiring fewer than

10^{5}

trainable parameters and only tens of MFLOPs per inference, enabling real-time operation on embedded and edge devices. Full article

(This article belongs to the Special Issue Signal Processing for Satellite Navigation and Wireless Localization)

► Show Figures

Figure 1

21 pages, 16524 KB

Open AccessArticle

MUSIC-Based Multi-Channel Forward-Scatter Radar Using OFDM Signals

by Yihua Qin, Abdollah Ajorloo and Fabiola Colone

Sensors 2025, 25(24), 7621; https://doi.org/10.3390/s25247621 - 16 Dec 2025

Abstract

This paper presents an advanced signal processing framework for multi-channel forward-scatter radar (MC-FSR) systems based on the Multiple Signal Classification (MUSIC) algorithm. The proposed framework addresses the inherent limitations of FFT-based space-domain processing, such as limited angular resolution and the poor detectability of [...] Read more.

This paper presents an advanced signal processing framework for multi-channel forward-scatter radar (MC-FSR) systems based on the Multiple Signal Classification (MUSIC) algorithm. The proposed framework addresses the inherent limitations of FFT-based space-domain processing, such as limited angular resolution and the poor detectability of weak or closely spaced targets, which become particularly severe in low-cost FSR systems, which are typically operated with small antenna arrays. The MUSIC algorithm is adapted to operate on real-valued data obtained from the non-coherent, amplitude-based MC-FSR approach by reformulating the steering vectors and adjusting the degrees of freedom (DoFs). While compatible with arbitrary transmitting waveforms, particular emphasis is placed on Orthogonal Frequency Division Multiplexing (OFDM) signals, which are widely used in modern communication systems such as Wi-Fi and cellular networks. An analysis of the OFDM waveform’s autocorrelation properties is provided to assess their impact on target detection, including strategies to mitigate rapid target signature decay using a sub-band approach and to manage signal correlation through spatial smoothing. Simulation results, including multi-target scenarios under constrained array configurations, demonstrate that the proposed MUSIC-based approach significantly enhances angular resolution and enables reliable discrimination of closely spaced targets even with a limited number of receiving channels. Experimental validation using an S-band MC-FSR prototype implemented with software-defined radios (SDRs) and commercial Wi-Fi antennas, involving cooperative targets like people and drones, further confirms the effectiveness and practicality of the proposed method for real-world applications. Overall, the proposed MUSIC-based MC-FSR framework exhibits strong potential for implementation in low-cost, hardware-constrained environments and is particularly suited for emerging Integrated Sensing and Communication (ISAC) systems. Full article

(This article belongs to the Special Issue Advances in Multichannel Radar Systems)

► Show Figures

Figure 1

Journal Description

Sensors

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Topical Collections

Further Information

Guidelines

MDPI Initiatives

Follow MDPI