Next Issue
Volume 25, September-1
Previous Issue
Volume 25, August-1
 
 
sensors-logo

Journal Browser

Journal Browser

Sensors, Volume 25, Issue 16 (August-2 2025) – 335 articles

Cover Story (view full-size image): The integration of robotics and next-generation mobile networks is reshaping telemedicine by enabling safe, remote interventions where specialists are scarce or exposed to risk. This paper introduces the IoRT-in-hand: a lightweight, smart end-effector that merges medical tools, vision, force sensing, and edge computing with Internet connectivity. Within an Edge–Cloud architecture, the system connects the physical robot to its Digital Twin, ensuring real-time visualization, safety, and control. Experiments confirm resilient performance in the presence of communication delays, supported by a modular design that integrates MQTT and ROS. This open-source approach promotes reproducibility and paves the way for future advances such as latency-aware control, autonomous assistance, and deployment in search-and-rescue or emergency scenarios. View this paper
  • Issues are regarded as officially published after their release is announced to the table of contents alert mailing list.
  • You may sign up for e-mail alerts to receive table of contents of newly released issues.
  • PDF is the official format for papers published in both, html and pdf forms. To view the papers in pdf format, click on the "PDF Full-text" link, and use the free Adobe Reader to open them.
Order results
Result details
Section
Select all
Export citation of selected articles as:
28 pages, 1036 KB  
Review
Recent Advances in Portable Dry Electrode EEG: Architecture and Applications in Brain-Computer Interfaces
by Meihong Zhang, Bocheng Qian, Jianming Gao, Shaokai Zhao, Yibo Cui, Zhiguo Luo, Kecheng Shi and Erwei Yin
Sensors 2025, 25(16), 5215; https://doi.org/10.3390/s25165215 - 21 Aug 2025
Viewed by 849
Abstract
As brain–computer interface (BCI) technology continues to advance, research on human brain function has gradually transitioned from theoretical investigation to practical engineering applications. To support EEG signal acquisition in a variety of real-world scenarios, BCI electrode systems must demonstrate a balanced combination of [...] Read more.
As brain–computer interface (BCI) technology continues to advance, research on human brain function has gradually transitioned from theoretical investigation to practical engineering applications. To support EEG signal acquisition in a variety of real-world scenarios, BCI electrode systems must demonstrate a balanced combination of electrical performance, wearing comfort, and portability. Dry electrodes have emerged as a promising alternative for EEG acquisition due to their ability to operate without conductive gel or complex skin preparation. This paper reviews the latest progress in dry electrode EEG systems, summarizing key achievements in hardware design with a focus on structural innovation and material development. It also examines application advances in several representative BCI domains, including emotion recognition, fatigue and drowsiness detection, motor imagery, and steady-state visual evoked potentials, while analyzing system-level performance. Finally, the paper critically assesses existing challenges and identifies critical future research priorities. Key recommendations include developing a standardized evaluation framework to bolster research reliability, enhancing generalization performance, and fostering coordinated hardware-algorithm optimization. These steps are crucial for advancing the practical implementation of these technologies across diverse scenarios. With this survey, we aim to offer a comprehensive reference and roadmap for researchers engaged in the development and implementation of next-generation dry electrode EEG-based BCI systems. Full article
Show Figures

Figure 1

36 pages, 7177 KB  
Article
Performance Optimization Analysis of Partial Discharge Detection Manipulator Based on STPSO-BP and CM-SA Algorithms
by Lisha Luo, Junjie Huang, Yuyuan Chen, Yujing Zhao, Jufang Hu and Chunru Xiong
Sensors 2025, 25(16), 5214; https://doi.org/10.3390/s25165214 - 21 Aug 2025
Viewed by 505
Abstract
In high-voltage switchgear, partial discharge (PD) detection using six-degree-of-freedom (6-DOF) manipulators presents challenges. However, these involve inverse kinematics (IK) solution redundancy and the lack of synergistic optimization between end-effector positioning accuracy and energy consumption. To address these issues, a dual-layer adaptive optimization model [...] Read more.
In high-voltage switchgear, partial discharge (PD) detection using six-degree-of-freedom (6-DOF) manipulators presents challenges. However, these involve inverse kinematics (IK) solution redundancy and the lack of synergistic optimization between end-effector positioning accuracy and energy consumption. To address these issues, a dual-layer adaptive optimization model integrating multiple algorithms is proposed. In the first layer, a spatio-temporal correlation particle memory-based particle swarm optimization BP neural network (STPSO-BP) is employed. It replaces traditional IK, while long short-term memory (LSTM) predicts particle movement trends, and trajectory similarity penalties constrain search trajectories. Thereby, positioning accuracy and adaptability are enhanced. In the second layer, a chaotic mapping-based simulated annealing (CM-SA) algorithm is utilized. Chaotic joint angle constraints, dynamic weight adjustment, and dynamic temperature regulation are incorporated. This approach achieves collaborative optimization of energy consumption and positioning error, utilizing cubic spline interpolation to smooth the joint trajectory. Specifically, the positioning error decreases by 68.9% compared with the traditional BP neural network algorithm. Energy consumption is reduced by 60.18% in contrast to the pre-optimization state. Overall, the model achieves significant optimization. An innovative solution for synergistic accuracy–energy control in 6-DOF manipulators for PD detection is offered. Full article
(This article belongs to the Section Sensors and Robotics)
Show Figures

Figure 1

17 pages, 891 KB  
Article
LLaVA-Pose: Keypoint-Integrated Instruction Tuning for Human Pose and Action Understanding
by Dewen Zhang, Tahir Hussain, Wangpeng An and Hayaru Shouno
Sensors 2025, 25(16), 5213; https://doi.org/10.3390/s25165213 - 21 Aug 2025
Viewed by 541
Abstract
Current vision–language models (VLMs) are well-adapted for general visual understanding tasks. However, they perform inadequately when handling complex visual tasks related to human poses and actions due to the lack of specialized vision–language instruction-following data. We introduce a method for generating such data [...] Read more.
Current vision–language models (VLMs) are well-adapted for general visual understanding tasks. However, they perform inadequately when handling complex visual tasks related to human poses and actions due to the lack of specialized vision–language instruction-following data. We introduce a method for generating such data by integrating human keypoints with traditional visual features such as captions and bounding boxes, enabling more precise understanding of human-centric scenes. Our approach constructs a dataset comprising 200,328 samples tailored to fine-tune models for human-centric tasks, focusing on three areas: conversation, detailed description, and complex reasoning. We establish an Extended Human Pose and Action Understanding Benchmark (E-HPAUB) to assess model performance on human pose and action understanding. We fine-tune the LLaVA-1.5-7B model using this dataset and evaluate our resulting LLaVA-Pose model on the benchmark, achieving significant improvements. Experimental results show an overall improvement of 33.2% compared to the original LLaVA-1.5-7B model. These findings highlight the effectiveness of keypoint-integrated data in enhancing multimodal models for human-centric visual understanding. Full article
(This article belongs to the Section Intelligent Sensors)
Show Figures

Figure 1

16 pages, 1430 KB  
Article
Assessing Smooth Pursuit Eye Movements Using Eye-Tracking Technology in Patients with Schizophrenia Under Treatment: A Pilot Study
by Luis Benigno Contreras-Chávez, Valdemar Emigdio Arce-Guevara, Luis Fernando Guerrero, Alfonso Alba, Miguel G. Ramírez-Elías, Edgar Roman Arce-Santana, Victor Hugo Mendez-Garcia, Jorge Jimenez-Cruz, Anna Maria Maddalena Bianchi and Martin O. Mendez
Sensors 2025, 25(16), 5212; https://doi.org/10.3390/s25165212 - 21 Aug 2025
Viewed by 659
Abstract
Schizophrenia is a complex disorder that affects mental organization and cognitive functions, including concentration and memory. One notable manifestation of cognitive changes in schizophrenia is a diminished ability to scan and perform tasks related to visual inspection. From the three evaluable aspects of [...] Read more.
Schizophrenia is a complex disorder that affects mental organization and cognitive functions, including concentration and memory. One notable manifestation of cognitive changes in schizophrenia is a diminished ability to scan and perform tasks related to visual inspection. From the three evaluable aspects of the ocular movements (saccadic, smooth pursuit, and fixation) in particular, smooth pursuit eye movement (SPEM) involves the tracking of slow moving objects and is closely related to attention, visual memory, and processing speed. However, evaluating smooth pursuit in clinical settings is challenging due to the technical complexities of detecting these movements, resulting in limited research and clinical application. This pilot study investigates whether the quantitative metrics derived from eye-tracking data can distinguish between patients with schizophrenia under treatment and healthy controls. The study included nine healthy participants and nine individuals receiving treatment for schizophrenia. Gaze trajectories were recorded using an eye tracker during a controlled visual tracking task performed during a clinical visit. Spatiotemporal analysis of gaze trajectories was performed by evaluating three different features: polygonal area, colocalities, and direction difference. Subsequently, a support vector machine (SVM) was used to assess the separability between healthy individuals and those with schizophrenia based on the identified gaze trajectory features. The results show statistically significant differences between the control and subjects with schizophrenia for all the computed indexes (p < 0.05) and a high separability achieving around 90% of accuracy, sensitivity, and specificity. The results suggest the potential development of a valuable clinical tool for the evaluation of SPEM, offering utility in clinics to assess the efficacy of therapeutic interventions in individuals with schizophrenia. Full article
(This article belongs to the Special Issue Biomedical Imaging, Sensing and Signal Processing)
Show Figures

Figure 1

21 pages, 39236 KB  
Article
Adaptive Image Deblurring Convolutional Neural Network with Meta-Tuning
by Quoc-Thien Ho, Minh-Thien Duong, Seongsoo Lee and Min-Cheol Hong
Sensors 2025, 25(16), 5211; https://doi.org/10.3390/s25165211 - 21 Aug 2025
Viewed by 573
Abstract
Motion blur is a complex phenomenon caused by the relative movement between an observed object and an imaging sensor during the exposure time, resulting in degradation in the image quality. Deep-learning-based methods, particularly convolutional neural networks (CNNs), have shown promise in motion deblurring. [...] Read more.
Motion blur is a complex phenomenon caused by the relative movement between an observed object and an imaging sensor during the exposure time, resulting in degradation in the image quality. Deep-learning-based methods, particularly convolutional neural networks (CNNs), have shown promise in motion deblurring. However, the small kernel sizes of CNNs limit their ability to achieve optimal performance. Moreover, supervised deep-learning-based deblurring methods often exhibit overfitting in their training datasets. Models trained on widely used synthetic blur datasets frequently fail to generalize in other blur domains in real-world scenarios and often produce undesired artifacts. To address these challenges, we propose the Spatial Feature Selection Network (SFSNet), which incorporates a Regional Feature Extractor (RFE) module to expand the receptive field and effectively select critical spatial features in order to improve the deblurring performance. In addition, we present the BlurMix dataset, which includes diverse blur types, as well as a meta-tuning strategy for effective blur domain adaptation. Our method enables the network to rapidly adapt to novel blur distributions with minimal additional training, and thereby improve generalization. The experimental results show that the meta-tuning variant of the SFSNet eliminates unwanted artifacts and significantly improves the deblurring performance across various blur domains. Full article
Show Figures

Figure 1

13 pages, 3255 KB  
Article
Application of the Composite Electrical Insulation Layer with a Self-Healing Function Similar to Pine Trees in K-Type Coaxial Thermocouples
by Zhenyin Hai, Yue Chen, Zhixuan Su, Hongwei Ji, Yihang Zhang, Shigui Gong, Shanmin Gao, Chenyang Xue, Libo Gao and Zhichun Liu
Sensors 2025, 25(16), 5210; https://doi.org/10.3390/s25165210 - 21 Aug 2025
Viewed by 516
Abstract
Aerospace engines and hypersonic vehicles, among other high-temperature components, often operate in environments characterized by temperatures exceeding 1000 °C and high-speed airflow impacts, resulting in severe thermal erosion conditions. Coaxial thermocouples (CTs), with their unique self-eroding characteristic, are particularly well suited for use [...] Read more.
Aerospace engines and hypersonic vehicles, among other high-temperature components, often operate in environments characterized by temperatures exceeding 1000 °C and high-speed airflow impacts, resulting in severe thermal erosion conditions. Coaxial thermocouples (CTs), with their unique self-eroding characteristic, are particularly well suited for use in such extreme environments. However, fabricating high-temperature electrical insulation layers for coaxial thermocouples remains challenging. Inspired by the self-healing mechanism of pine trees, we designed a composite electrical insulation layer with a similar self-healing function. This composite layer exhibits excellent high-temperature insulation properties (insulation resistance of 14.5 kΩ at 1200 °C). Applied as the insulation layer in K-type coaxial thermocouples via dip-coating, the thermocouples were tested for temperature and heat flux. Temperature tests showed an accuracy of 1.72% in the range of 200–1200 °C, a drift rate better than 0.474%/h at 1200 °C, and hysteresis better than 0.246%. The temperature response time was 1.08 ms. Heat flux tests demonstrated a measurable range of 0–41.32 MW/m2 with an accuracy better than 6.511% and a heat flux response time of 7.6 ms. In simulated extreme environments, the K-type coaxial thermocouple withstood 70 s of 900 °C flame impact and 50 cycles of high-power laser thermal shock. Full article
(This article belongs to the Special Issue Advancements and Applications of Biomimetic Sensors Technologies)
Show Figures

Figure 1

26 pages, 1971 KB  
Article
Dynamic Allocation of C-V2X Communication Resources Based on Graph Attention Network and Deep Reinforcement Learning
by Zhijuan Li, Guohong Li, Zhuofei Wu, Wei Zhang and Alessandro Bazzi
Sensors 2025, 25(16), 5209; https://doi.org/10.3390/s25165209 - 21 Aug 2025
Viewed by 507
Abstract
Vehicle-to-vehicle (V2V) and vehicle-to-network (V2N) communications are two key components of intelligent transport systems (ITSs) that can share spectrum resources through in-band overlay. V2V communication primarily supports traffic safety, whereas V2N primarily focuses on infotainment and information exchange. Achieving reliable V2V transmission alongside [...] Read more.
Vehicle-to-vehicle (V2V) and vehicle-to-network (V2N) communications are two key components of intelligent transport systems (ITSs) that can share spectrum resources through in-band overlay. V2V communication primarily supports traffic safety, whereas V2N primarily focuses on infotainment and information exchange. Achieving reliable V2V transmission alongside high-rate V2N services in resource-constrained, dynamically changing traffic environments poses a significant challenge for resource allocation. To address this, we propose a novel reinforcement learning (RL) framework, termed Graph Attention Network (GAT)-Advantage Actor–Critic (GAT-A2C). In this framework, we construct a graph based on V2V links and their potential interference relationships. Each V2V link is represented as a node, and edges connect nodes that may interfere. The GAT captures key interference patterns among neighboring vehicles while accounting for real-time mobility and channel variations. The features generated by the GAT, combined with individual link characteristics, form the environment state, which is then processed by the RL agent to jointly optimize the resource blocks allocation and the transmission power for both V2V and V2N communications. Simulation results demonstrate that the proposed method substantially improves V2N rates and V2V communication success ratios under various vehicle densities. Furthermore, the approach exhibits strong scalability, making it a promising solution for future large-scale intelligent vehicular networks operating in dynamic traffic scenarios. Full article
Show Figures

Figure 1

18 pages, 8907 KB  
Article
Arc Dynamics and Erosion Behavior of Pantograph-Catenary Contacts Under Controlled Humidity Levels
by Bingquan Li, Yijian Zhao, Ran Ji, Huajun Dong and Ningning Wei
Sensors 2025, 25(16), 5208; https://doi.org/10.3390/s25165208 - 21 Aug 2025
Viewed by 500
Abstract
In response to the instability fluctuations and erosion characteristic changes in pantograph-catenary system (PCS) arcs induced by humidity variations in an open environment, a single-variable controlled experimental approach based on multi-source data fusion is proposed. This study innovatively establishes a humidity-controlled reciprocating current-carrying [...] Read more.
In response to the instability fluctuations and erosion characteristic changes in pantograph-catenary system (PCS) arcs induced by humidity variations in an open environment, a single-variable controlled experimental approach based on multi-source data fusion is proposed. This study innovatively establishes a humidity-controlled reciprocating current-carrying arc initiation test platform, integrating digital image processing with the dynamic analysis of multi-physics sensor signals (current, voltage, temperature). The study quantitatively evaluates the arc motion characteristics and the erosion effects on the frictional contact pair under different relative humidity levels (30%, 50%, 70%, and 90%) with a DC power supply (120 V/25 A). The experimental data and analysis reveal that increasing humidity results in higher contact resistance and accumulated arc energy, with arc stability first improving and then decreasing. At low humidity, arc behavior is more intense, and the erosion rate is faster. As humidity increases, the electrode wear transitions from adhesive wear to electrochemical wear, accompanied by copper transfer. The results suggest that the arc stability is optimal at moderate humidity (50% RH), with a peak current-carrying efficiency of 66% and a minimum loss rate of 14.5%. This threshold offers a vital theoretical framework for the optimization and risk assessments of PCS design. Full article
(This article belongs to the Section Electronic Sensors)
Show Figures

Figure 1

17 pages, 1723 KB  
Article
HoneyLite: A Lightweight Honeypot Security Solution for SMEs
by Nurayn AlQahtan, Aseel AlOlayan, AbdulAziz AlAjaji and Abdulaziz Almaslukh
Sensors 2025, 25(16), 5207; https://doi.org/10.3390/s25165207 - 21 Aug 2025
Viewed by 464
Abstract
Small and medium-sized enterprises (SMEs) are increasingly targeted by cyber threats but often lack the financial and technical resources to implement advanced security systems. This paper presents HoneyLite, a lightweight and dynamic honeypot-based security solution specifically designed to meet the constraints and cybersecurity [...] Read more.
Small and medium-sized enterprises (SMEs) are increasingly targeted by cyber threats but often lack the financial and technical resources to implement advanced security systems. This paper presents HoneyLite, a lightweight and dynamic honeypot-based security solution specifically designed to meet the constraints and cybersecurity needs of SMEs. Unlike traditional honeypots, HoneyLite integrates real-time network traffic analysis with automated malware detection via the VirusTotal API, enabling it to identify a wide range of cyber threats, including TCP scans, FTP/SSH intrusions, ICMP flood attacks, and malicious file uploads. Developed using open-source tools, the system operates with minimal resource overhead and is validated within a simulated virtual environment. It also generates detailed threat reports to support incident analysis and response. By combining affordability, adaptability, and comprehensive threat visibility, HoneyLite offers a practical and scalable solution to help SMEs detect, analyze, and respond to modern cyberattacks in real time. Full article
(This article belongs to the Special Issue IoT Network Security (Second Edition))
Show Figures

Figure 1

17 pages, 2708 KB  
Article
Simulation and Implementation of the Modeling of Forklift with Tricycle in Warehouse Systems for ROS
by Kuo-Yang Tu, Che-Ping Hung, Hong-Yu Lin and Kaun-Yu Lin
Sensors 2025, 25(16), 5206; https://doi.org/10.3390/s25165206 - 21 Aug 2025
Viewed by 461
Abstract
In the age of labor shortage, increasing the throughput of warehouses is a good issue. In the recent two decades, automatic warehouses designed to reduce human labor have therefore become a very hot research topic. Tricycle forklifts being able to carry heavy goods [...] Read more.
In the age of labor shortage, increasing the throughput of warehouses is a good issue. In the recent two decades, automatic warehouses designed to reduce human labor have therefore become a very hot research topic. Tricycle forklifts being able to carry heavy goods can play important roles in automatic warehouses. Meanwhile, Robot Operating System (ROS) is a very famous and popular platform for developing the software of robotics. Its powerful communication function makes lots of warehouse information exchange easy. Therefore, ROS installed as the communication backbone of warehouse is very popular. However, the software modules of ROS do not offer tricycle forklifts. Therefore, in this research, the model of a tricycle forklift developed for ROS systems in warehouse applications is constructed. In spite of the developed model, the existing software modules must be modified for compatible connection such that the tricycle forklift can be navigated and controlled by constructed ROS. For the function of Simultaneous Localization And Mapping (SLAM) and the control of self-guided navigation, the constructed system is verified by Gazebo simulation. In addition, the experiments of a real tricycle forklift to demonstrate the developed ROS for enough accuracy of warehouse application are also included. Full article
(This article belongs to the Special Issue New Challenges and Sensor Techniques in Robot Positioning)
Show Figures

Figure 1

19 pages, 4825 KB  
Article
Design of a Novel Electromagnetic Ultrasonic Transducer for Stress Detection
by Changhong Chen, Chunguang Xu, Guangcan Yang, Yongjiang Ma and Shuangxu Yang
Sensors 2025, 25(16), 5205; https://doi.org/10.3390/s25165205 - 21 Aug 2025
Viewed by 554
Abstract
Accurate stress evaluation of structural components during manufacturing and operation is essential for ensuring the safety and reliability of advanced equipment in aerospace, defense, and other high-performance fields. However, existing electromagnetic ultrasonic stress detection methods are often limited by low signal amplitude and [...] Read more.
Accurate stress evaluation of structural components during manufacturing and operation is essential for ensuring the safety and reliability of advanced equipment in aerospace, defense, and other high-performance fields. However, existing electromagnetic ultrasonic stress detection methods are often limited by low signal amplitude and limited adaptability to complex environments, hindering their practical deployment for in situ testing. This study proposes a novel surface wave transducer structure for stress detection based on acoustoelastic theory combined with electromagnetic ultrasonic technology. It innovatively designs a surface wave transducer composed of multiple proportionally scaled dislocation meandering coils. This innovative configuration significantly enhances the Lorentz force distribution and coupling efficiency, which accurately measure the stress of components through acoustic time delays and present an experimental method for applying electromagnetic ultrasonic technology to in situ stress detection. Finite element simulations confirmed the optimized acoustic field characteristics, and experimental validation on 6061 aluminum alloy specimens demonstrated a 111.1% improvement in signal amplitude compared to conventional designs. Through multiple experiments and curve fitting, the average relative error of the measurement results is less than 4.53%, verifying the accuracy of the detection method. Further testing under random stress conditions validated the transducer’s feasibility for in situ testing in production and service environments. Owing to its enhanced signal strength, compact structure, and suitability for integration with automated inspection systems, the proposed transducer shows strong potential for in situ stress monitoring in demanding industrial environments. Full article
(This article belongs to the Section Fault Diagnosis & Sensors)
Show Figures

Figure 1

19 pages, 1904 KB  
Article
Feasibility of Wearable Devices for Motivating Post-Stroke Patients
by Klaudia Marek, Jan Górski, Piotr Karolczyk, Justyna Redlicka, Igor Zubrycki and Elżbieta Miller
Sensors 2025, 25(16), 5204; https://doi.org/10.3390/s25165204 - 21 Aug 2025
Viewed by 660
Abstract
The effectiveness of upper extremity rehabilitation in post-stroke patients significantly depends on patient motivation and adherence to therapeutic regimens. Rehabilitation-assistive technologies, including wearable sensors, have been adopted to facilitate intensive and repetitive exercises aimed at reducing hand dysfunction and enhancing quality of life. [...] Read more.
The effectiveness of upper extremity rehabilitation in post-stroke patients significantly depends on patient motivation and adherence to therapeutic regimens. Rehabilitation-assistive technologies, including wearable sensors, have been adopted to facilitate intensive and repetitive exercises aimed at reducing hand dysfunction and enhancing quality of life. Building upon the previously introduced Przypominajka (reminder) system reported in this journal—a wearable sensory glove coupled with a mobile application providing exercise guidance and monitoring—we conducted a feasibility study to evaluate its effectiveness in supporting upper limb rehabilitation. Sixteen post-stroke patients with hemiparesis were equally randomized into experimental and control groups. Both groups performed upper limb exercises for 45 min daily for over two weeks. The experimental group utilized the sensor-equipped glove and tablet-based exercises, whereas the control group followed printed exercise instructions. Clinical improvements were measured using the Fugl–Meyer Assessment–Upper Extremity (FMA-UE), Functional Independence Measure (FIM), and MORE scales. The experimental group demonstrated a minimal clinically important difference (MCID) on the FMA-UE and reported greater overall improvement than the control group. This study confirms the feasibility and potential clinical benefit of supplementing post-stroke rehabilitation with sensor-augmented exercises provided by the previously described Przypominajka device. Full article
(This article belongs to the Special Issue Sensors-Based Healthcare Diagnostics, Monitoring and Medical Devices)
Show Figures

Figure 1

59 pages, 3591 KB  
Review
Efficient Caching Strategies in NDN-Enabled IoT Networks: Strategies, Constraints, and Future Directions
by Ala’ Ahmad Alahmad, Azana Hafizah Mohd Aman, Faizan Qamar and Wail Mardini
Sensors 2025, 25(16), 5203; https://doi.org/10.3390/s25165203 - 21 Aug 2025
Viewed by 450
Abstract
Named Data Networking (NDN) is identified as a significant shift within the information-centric networking (ICN) perspective that avoids our current IP-based infrastructures by retrieving data based on its name rather than where the host is placed. This shift in paradigm is especially beneficial [...] Read more.
Named Data Networking (NDN) is identified as a significant shift within the information-centric networking (ICN) perspective that avoids our current IP-based infrastructures by retrieving data based on its name rather than where the host is placed. This shift in paradigm is especially beneficial in Internet of Things (IoT) settings because information sharing is a critical challenge, as millions of IoT items create enormous traffic. Content caching in the network is another key characteristic of NDN used in IoT, which enables data storing within the network and provides IoT devices with the opportunity to address nearby caching nodes to gain the intended content, which, in its turn, will minimize latency as well as bandwidth consumption. However, effective caching solutions must be developed since cache management is made difficult by the constant shifting of IoT networks and the constrained capabilities of IoT devices. This paper gives an overview of cache strategies in NDN-based IoT systems. It emphasizes six strategy types: popularity-based, freshness-aware, collaborative, hybrid, probabilistic, and machine learning-based, evaluating their performances in terms of demands like content preference, cache update, and power consumption. By analyzing various caching policies and their performance characteristics, this paper provides valuable insights for researchers and practitioners developing caching strategies in NDN-based IoT networks. Full article
(This article belongs to the Section Internet of Things)
Show Figures

Figure 1

23 pages, 3538 KB  
Article
VCformer: Variable-Centric Multi-Scale Transformer for Multivariate Time Series Forecasting
by Junyu Zhu, Enguang Zuo, Xinyu Bi, Chen Chen, Cheng Chen, Ziwei Yan and Xiaoyi Lv
Sensors 2025, 25(16), 5202; https://doi.org/10.3390/s25165202 - 21 Aug 2025
Viewed by 613
Abstract
Multivariate time series forecasting is crucial for numerous practical applications ranging from financial markets to climate monitoring. Traditional multivariate time series forecasting methods primarily adopt a time-centric modeling paradigm, applying attention mechanisms to the temporal dimension, which presents significant limitations when handling complex [...] Read more.
Multivariate time series forecasting is crucial for numerous practical applications ranging from financial markets to climate monitoring. Traditional multivariate time series forecasting methods primarily adopt a time-centric modeling paradigm, applying attention mechanisms to the temporal dimension, which presents significant limitations when handling complex dependencies between variables. To better capture inter-variable interaction patterns, this paper proposes the Variable-Centric Transformer (VCformer), which shifts the attention paradigm from time-centric to variable-centric through sequence transposition. Building upon this foundation, we further design a dual-scale architecture that simultaneously models feature representations at both the original variable level and variable group level. Combined with an adaptive variable grouping mechanism, we construct a parameter-sharing dual-path encoder and finally select the optimal feature fusion strategy through comparative experiments. Experimental results on seven benchmark datasets demonstrate that VCformer achieves comprehensive improvements in prediction accuracy compared to traditional time-centric methods, while exhibiting stronger modeling capabilities on high-dimensional data. Ablation studies and interpretability analysis further validate the effectiveness of each component. Full article
(This article belongs to the Section Intelligent Sensors)
Show Figures

Figure 1

34 pages, 1151 KB  
Article
Innovative Technologies to Improve Occupational Safety in Mining and Construction Industries—Part I
by Paweł Bęś, Paweł Strzałkowski, Justyna Górniak-Zimroz, Mariusz Szóstak and Mateusz Janiszewski
Sensors 2025, 25(16), 5201; https://doi.org/10.3390/s25165201 - 21 Aug 2025
Viewed by 824
Abstract
Innovative technologies have been helping to improve comfort and safety at work in high-risk sectors for years. The study analysed the impact, along with an assessment of potential implementations (opportunities and limitations) of innovative technological solutions for improving occupational safety in two selected [...] Read more.
Innovative technologies have been helping to improve comfort and safety at work in high-risk sectors for years. The study analysed the impact, along with an assessment of potential implementations (opportunities and limitations) of innovative technological solutions for improving occupational safety in two selected sectors of the economy: mining and construction. The technologies evaluated included unmanned aerial vehicles and inspection robots, the Internet of Things and sensors, artificial intelligence, virtual and augmented reality, innovative individual and collective protective equipment, and exoskeletons. Due to the extensive nature of the obtained materials, the research description has been divided into two articles (Part I and Part II). This article presents the first three technologies. After the scientific literature from the Scopus database was analysed, some research gaps that need to be filled were identified. In addition to the obvious benefits of increased occupational safety for workers, innovative technological solutions also offer employers several economic advantages that affect the industry’s sustainability. Innovative technologies are playing an increasingly important role in improving safety in mining and construction. However, further integration and overcoming implementation barriers, such as the need for changes in education, are needed to realise their full potential. Full article
(This article belongs to the Section Industrial Sensors)
Show Figures

Figure 1

20 pages, 10705 KB  
Article
EMFE-YOLO: A Lightweight Small Object Detection Model for UAVs
by Chengjun Yang, Yan Shen and Lutao Wang
Sensors 2025, 25(16), 5200; https://doi.org/10.3390/s25165200 - 21 Aug 2025
Viewed by 792
Abstract
Small object detection in Unmanned Aerial Vehicles’ (UAVs) aerial images faces challenges such as low detection accuracy and complex backgrounds. Meanwhile, it is difficult to deploy the object detection models with large parameters on resource-constrained UAVs. Therefore, a lightweight small object detection model [...] Read more.
Small object detection in Unmanned Aerial Vehicles’ (UAVs) aerial images faces challenges such as low detection accuracy and complex backgrounds. Meanwhile, it is difficult to deploy the object detection models with large parameters on resource-constrained UAVs. Therefore, a lightweight small object detection model EMFE-YOLO is proposed based on efficient multi-scale feature enhancement by improving YOLOv8s. Firstly, the Enhanced Attention to Large-scale Features (EALF) structure is applied in EMFE-YOLO to focus on large-scale features, improve the detection ability to small objects, and decrease the parameters. Secondly, the efficient multi-scale feature enhancement (EMFE) module is integrated into the backbone of EALF for feature extraction and enhancement. The EMFE module reduces the computational cost, obtains richer contextual information, and mitigates the interference from complex backgrounds. Finally, DySample is employed in the neck of EALF to optimize the upsampling process of features. The EMFE-YOLO is validated on the VisDrone2019-val dataset. Experimental results show that it improves mAP50 and mAP50:95 by 8.5% and 6.3%, respectively, and reduces the parameters by 73% compared to YOLOv8s. These results demonstrate that EMFE-YOLO achieves a good balance between accuracy and efficiency, making it suitable for deployment on UAVs with limited computational resources. Full article
(This article belongs to the Section Sensing and Imaging)
Show Figures

Figure 1

16 pages, 5540 KB  
Article
Sensor-Driven RSSI Prediction via Adaptive Machine Learning and Environmental Sensing
by Anya Apavatjrut
Sensors 2025, 25(16), 5199; https://doi.org/10.3390/s25165199 - 21 Aug 2025
Viewed by 500
Abstract
Received Signal Strength Indicator (RSSI) prediction is valuable for network planning and optimization as it helps determine the optimal placements of wireless access points and enables better coverage planning. It is also crucial for efficient handover management between cells or access points, reducing [...] Read more.
Received Signal Strength Indicator (RSSI) prediction is valuable for network planning and optimization as it helps determine the optimal placements of wireless access points and enables better coverage planning. It is also crucial for efficient handover management between cells or access points, reducing dropped connections and improving service quality. Additionally, RSSI prediction supports indoor positioning systems, power management optimization, and cost-efficient network deployment. Path loss models have historically served as the foundation for RSSI prediction, providing a theoretical framework for estimating signal strength degradation. However, modern machine learning approaches have emerged as a revolutionary solution for network optimization, providing more versatile and data-driven methods to enhance wireless network performance. In this paper, an adaptive machine learning framework integrating environmental sensing parameters such as temperature, relative humidity, barometric pressure, and particulate matter for RSSI prediction is proposed. Performance analysis reveals that RSSI values are influenced by environmental factors through complex, non-linear interactions, thereby challenging the conventional linear assumptions of traditional path loss models. The proposed model demonstrates improved predictive accuracy over the baseline, with relative increases in variance explained of 6.02% and 2.04% compared to the baseline model excluding and including environmental parameters, respectively. Additionally, the root mean squared error is reduced to 1.40 dB. These results demonstrate that cognitive methods incorporating environmental data can substantially enhance RSSI prediction accuracy in wireless communications. Full article
Show Figures

Figure 1

15 pages, 5506 KB  
Article
Polyimide-Based Flexible Microelectrode Array for Non-Invasive Transcorneal Electrical Stimulation
by Víctor Manuel Carpio-Verdín, Natiely Hernández-Sebastián, Bernardino Barrientos-García, Silvia Solis-Ortiz, Erik R. Bojorges-Valdez, Francisco López-Huerta, Carlos Ismael Mares-Castro and Wilfrido Calleja-Arriaga
Sensors 2025, 25(16), 5198; https://doi.org/10.3390/s25165198 - 21 Aug 2025
Viewed by 876
Abstract
Transcorneal electrical stimulation (TES) is a promising treatment for several retinal degenerative diseases (RDDs). TES involves the application of a controlled electrical current to the anterior surface of the cornea, aimed at activating the retina and posterior ocular structures. Dawson–Trick–Litzkow (DTL) and ERG-JET [...] Read more.
Transcorneal electrical stimulation (TES) is a promising treatment for several retinal degenerative diseases (RDDs). TES involves the application of a controlled electrical current to the anterior surface of the cornea, aimed at activating the retina and posterior ocular structures. Dawson–Trick–Litzkow (DTL) and ERG-JET electrodes are among the most widely used for TES. However, their continuous metallic surface design limits spatial resolution and the ability to perform selective ES. In this work, we present the development of a transcorneal electrical stimulation (TES) electrode that, unlike conventional electrodes, enables spatially selective TES. The proposed electrode design consists of an array of 20 independent microelectrodes distributed across the central and paracentral regions of the cornea. The fabrication process combines surface micromachining and flexible electronics technologies, employing only three structural materials: aluminum (Al), titanium (Ti), and polyimide (PI). This material selection is critical for achieving a simplified, reproducible, and low-cost fabrication process. The fabricated electrode was validated through electrical and electrochemical testing. The results show a relatively high electrical conductivity of Al/Ti structures, low electrochemical impedance values—ranging from 791 kΩ to 1.75 MΩ for the clinically relevant frequency range (11 to 30 Hz)—and a high charge storage capacity of 1437 mC/cm2. The electrode capacity for electrical signal transmission was demonstrated through in vitro testing. Finally, the applicability of the TES electrode for electroretinogram (ERG) recording was evaluated by measuring its optical transmittance across the visible wavelength range. Full article
Show Figures

Figure 1

23 pages, 17970 KB  
Article
Strain Monitoring and Numerical Simulation Analysis of Nuclear Containment Structure During Containment Tests
by Xunqiang Yin, Weilong Yang, Junkai Zhang, Min Zhao and Jianbo Li
Sensors 2025, 25(16), 5197; https://doi.org/10.3390/s25165197 - 21 Aug 2025
Viewed by 424
Abstract
Strain monitoring during the service life of a nuclear containment structure is an effective means to evaluate whether the structure is operating safely. Due to the failure of embedded strain sensors, surface-mounted strain sensors should be installed on the outer wall of the [...] Read more.
Strain monitoring during the service life of a nuclear containment structure is an effective means to evaluate whether the structure is operating safely. Due to the failure of embedded strain sensors, surface-mounted strain sensors should be installed on the outer wall of the structure. However, whether the data from these substitute sensors can reasonably reflect the internal deformation behavior requires further investigation. To ensure the feasibility of the added strain sensors, a refined 3D model of a Chinese Pressurized Reactor (CPR1000) nuclear containment structure was developed in ANSYS 19.1 to study the internal and external deformation laws during a containment test (CTT). Solid reinforcement and cooling methods were employed to simulate prestressed cables and pre-tension application. The influence of ordinary steel bars in concrete was modeled using the smeared model, while interactions between the steel liner and concrete were simulated through coupled nodes. The model’s validity was verified against embedded strain sensor data recorded during a CTT. Furthermore, concrete and prestressed material parameters were refined through a sensitivity analysis. Finally, the variation law between the internal and external deformation of the containment structure was investigated under typical CTT loading conditions. Strain values in the wall thickness direction exhibited an essentially linear relationship. Near the equipment hatch, however, the strain distribution pattern was significantly influenced by the spatial arrangement of prestressed cables. Refined FEM and sensor systems are vital containment monitoring tools. Critically, surface-mounted strain sensors offer a feasible approach for inferring internal stress states and deformation behavior. This study provides theoretical support and a technical foundation for the safe assessment and maintenance of nuclear containment structures during operational service. Full article
(This article belongs to the Section Physical Sensors)
Show Figures

Figure 1

22 pages, 6186 KB  
Article
Introducing Fast Fourier Convolutions into Anomaly Detection
by Zhen Zhao and Jiali Zhou
Sensors 2025, 25(16), 5196; https://doi.org/10.3390/s25165196 - 21 Aug 2025
Viewed by 590
Abstract
Anomaly detection is inherently challenging, as anomalies typically emerge only at test time. While reconstruction-based methods are popular, their reliance on CNN backbones with local receptive fields limits discrimination and precise localization. We propose FFC-AD, a reconstruction framework using Fourier Feature Convolutions (FFCs) [...] Read more.
Anomaly detection is inherently challenging, as anomalies typically emerge only at test time. While reconstruction-based methods are popular, their reliance on CNN backbones with local receptive fields limits discrimination and precise localization. We propose FFC-AD, a reconstruction framework using Fourier Feature Convolutions (FFCs) to capture global information early, and we introduce Hidden Space Anomaly Simulation (HSAS), a latent-space regularization strategy that mitigates overgeneralization. Experiments on MVTec AD and VisA demonstrate that FFC-AD significantly outperforms state-of-the-art methods in both detection and segmentation accuracy. Full article
(This article belongs to the Section Sensor Networks)
Show Figures

Figure 1

17 pages, 2418 KB  
Article
InstructSee: Instruction-Aware and Feedback-Driven Multimodal Retrieval with Dynamic Query Generation
by Guihe Gu, Yuan Xue, Zhengqian Wu, Lin Song and Chao Liang
Sensors 2025, 25(16), 5195; https://doi.org/10.3390/s25165195 - 21 Aug 2025
Viewed by 492
Abstract
In recent years, cross-modal retrieval has garnered significant attention due to its potential to bridge heterogeneous data modalities, particularly in aligning visual content with natural language. Despite notable progress, existing methods often struggle to accurately capture user intent when queries are expressed through [...] Read more.
In recent years, cross-modal retrieval has garnered significant attention due to its potential to bridge heterogeneous data modalities, particularly in aligning visual content with natural language. Despite notable progress, existing methods often struggle to accurately capture user intent when queries are expressed through complex or evolving instructions. To address this challenge, we propose a novel cross-modal representation learning framework that incorporates an instruction-aware dynamic query generation mechanism, augmented by the semantic reasoning capabilities of large language models (LLMs). The framework dynamically constructs and iteratively refines query representations conditioned on natural language instructions and guided by user feedback, thereby enabling the system to effectively infer and adapt to implicit retrieval intent. Extensive experiments on standard multimodal retrieval benchmarks demonstrate that our method significantly improves retrieval accuracy and adaptability, outperforming fixed-query baselines and showing enhanced cross-modal alignment and generalization across diverse retrieval tasks. Full article
Show Figures

Figure 1

23 pages, 6924 KB  
Article
A Dynamic Multi-Scale Feature Fusion Network for Enhanced SAR Ship Detection
by Rui Cao and Jianghua Sui
Sensors 2025, 25(16), 5194; https://doi.org/10.3390/s25165194 - 21 Aug 2025
Viewed by 563
Abstract
This study aims to develop an enhanced YOLO algorithm to improve the ship detection performance of synthetic aperture radar (SAR) in complex marine environments. Current SAR ship detection methods face numerous challenges in complex sea conditions, including environmental interference, false detection, and multi-scale [...] Read more.
This study aims to develop an enhanced YOLO algorithm to improve the ship detection performance of synthetic aperture radar (SAR) in complex marine environments. Current SAR ship detection methods face numerous challenges in complex sea conditions, including environmental interference, false detection, and multi-scale changes in detection targets. To address these issues, this study adopts a technical solution that combines multi-level feature fusion with a dynamic detection mechanism. First, a cross-stage partial dynamic channel transformer module (CSP_DTB) was designed, which combines the transformer architecture with a convolutional neural network to replace the last two C3k2 layers in the YOLOv11n main network, thereby enhancing the model’s feature extraction capabilities. Second, a general dynamic feature pyramid network (RepGFPN) was introduced to reconstruct the neck network architecture, enabling more efficient multi-scale feature fusion and information propagation. Additionally, a lightweight dynamic decoupled dual-alignment head (DYDDH) was constructed to enhance the collaborative performance of localization and classification tasks through task-specific feature decoupling. Experimental results show that the proposed DRGD-YOLO algorithm achieves significant performance improvements. On the HRSID dataset, the algorithm achieves an average precision (mAP50) of 93.1% at an IoU threshold of 0.50 and an mAP50–95 of 69.2% over the IoU threshold range of 0.50–0.95. Compared to the baseline YOLOv11n algorithm, the proposed method improves mAP50 and mAP50–95 by 3.3% and 4.6%, respectively. The proposed DRGD-YOLO algorithm not only significantly improves the accuracy and robustness of synthetic aperture radar (SAR) ship detection but also demonstrates broad application potential in fields such as maritime surveillance, fisheries management, and maritime safety monitoring, providing technical support for the development of intelligent marine monitoring technology. Full article
(This article belongs to the Section Navigation and Positioning)
Show Figures

Figure 1

17 pages, 2751 KB  
Article
Joint Extraction of Cyber Threat Intelligence Entity Relationships Based on a Parallel Ensemble Prediction Model
by Huan Wang, Shenao Zhang, Zhe Wang, Jing Sun and Qingzheng Liu
Sensors 2025, 25(16), 5193; https://doi.org/10.3390/s25165193 - 21 Aug 2025
Viewed by 514
Abstract
The construction of knowledge graphs in cyber threat intelligence (CTI) critically relies on automated entity–relation extraction. However, sequence tagging-based methods for joint entity–relation extraction are affected by the order-dependency problem. As a result, overlapping relations are handled ineffectively. To address this limitation, a [...] Read more.
The construction of knowledge graphs in cyber threat intelligence (CTI) critically relies on automated entity–relation extraction. However, sequence tagging-based methods for joint entity–relation extraction are affected by the order-dependency problem. As a result, overlapping relations are handled ineffectively. To address this limitation, a parallel, ensemble-prediction–based model is proposed for joint entity–relation extraction in CTI. The joint extraction task is reformulated as an ensemble prediction problem. A joint network that combines Bidirectional Encoder Representations from Transformers (BERT) with a Bidirectional Gated Recurrent Unit (BiGRU) is constructed to capture deep contextual features in sentences. An ensemble prediction module and a triad representation of entity–relation facts are designed for joint extraction. A non-autoregressive decoder is employed to generate relation triad sets in parallel, thereby avoiding unnecessary sequential constraints during decoding. In the threat intelligence domain, labeled data are scarce and manual annotation is costly. To mitigate these constraints, the SecCti dataset is constructed by leveraging ChatGPT’s small-sample learning capability for labeling and augmentation. This approach reduces annotation costs effectively. Experimental results show a 4.6% absolute F1 improvement over the baseline on joint entity–relation extraction for threat intelligence concerning Advanced Persistent Threats (APTs) and cybercrime activities. Full article
(This article belongs to the Section Sensor Networks)
Show Figures

Figure 1

19 pages, 7241 KB  
Article
RICNET: Retinex-Inspired Illumination Curve Estimation for Low-Light Enhancement in Industrial Welding Scenes
by Chenbo Shi, Xiangyu Zhang, Delin Wang, Changsheng Zhu, Aiping Liu, Chun Zhang and Xiaobing Feng
Sensors 2025, 25(16), 5192; https://doi.org/10.3390/s25165192 - 21 Aug 2025
Viewed by 462
Abstract
Feature tracking is essential for welding crawler robots’ trajectory planning. As welding often occurs in dark environments like pipelines or ship hulls, the system requires low-light image capture for laser tracking. However, such images typically have poor brightness and contrast, degrading both weld [...] Read more.
Feature tracking is essential for welding crawler robots’ trajectory planning. As welding often occurs in dark environments like pipelines or ship hulls, the system requires low-light image capture for laser tracking. However, such images typically have poor brightness and contrast, degrading both weld seam feature extraction and trajectory anomaly detection accuracy. To address this, we propose a Retinex-based low-light enhancement network tailored for cladding scenarios. The network features an illumination curve estimation module and requires no paired or unpaired reference images during training, alleviating the need for cladding-specific datasets. It adaptively adjusts brightness, restores image details, and effectively suppresses noise. Extensive experiments on public (LOLv1 and LOLv2) and self-collected weld datasets show that our method outperformed existing approaches in PSNR, SSIM, and LPIPS. Additionally, weld seam segmentation under low-light conditions achieved 95.1% IoU and 98.9% accuracy, confirming the method’s effectiveness for downstream tasks in robotic welding. Full article
(This article belongs to the Section Optical Sensors)
Show Figures

Figure 1

17 pages, 5184 KB  
Article
Preparation and Applications of Silver Nanowire-Polyurethane Flexible Sensor
by Jiangyin Shan, Jianhua Qian, Ling Lin, Mengrong Wei, Jingyue Xia and Lin Fu
Sensors 2025, 25(16), 5191; https://doi.org/10.3390/s25165191 - 21 Aug 2025
Viewed by 578
Abstract
To expand the application of silver nanowires (AgNWs) in the field of flexible sensors, this study developed a stretchable flexible sensor based on thermoplastic polyurethane (TPU). Initially, the TPU nanofiber membrane was prepared by electrospinning. Subsequently, high-aspect-ratio AgNWs were synthesized via a one-step [...] Read more.
To expand the application of silver nanowires (AgNWs) in the field of flexible sensors, this study developed a stretchable flexible sensor based on thermoplastic polyurethane (TPU). Initially, the TPU nanofiber membrane was prepared by electrospinning. Subsequently, high-aspect-ratio AgNWs were synthesized via a one-step polyol reduction method. The AgNWs with the optimal aspect ratio were selected for the conductive layer and spray-coated onto the surface of the TPU nanofiber membrane. Another layer of TPU nanofiber membrane was then laminated on top, resulting in a flexible thin-film sensor with a “sandwich” structure. Through morphological, chemical structure, and crystallinity analyses, the primary factors influencing AgNWs’ growth were investigated. Performance tests revealed that the prepared AgNWs had an average length of approximately 130 μm, a diameter of about 80 nm, and an average aspect ratio exceeding 1500, with the highest being 1921. The obtained sensor exhibited a low initial resistance (26.7 Ω), high strain range (sensing, ε = 0–150%), high sensitivity (GF, over 19.21), fast response and recovery time (112 ms), and excellent conductivity (428 S/cm). Additionally, the sensor maintained stable resistance after 3000 stretching cycles at a strain range of 0–10%. The sensor could output stable and recognizable electrical signals, demonstrating significant potential for applications in motion monitoring, human–computer interaction, and healthcare fields. Full article
Show Figures

Graphical abstract

26 pages, 8663 KB  
Article
Acoustics-Augmented Diagnosis Method for Rolling Bearings Based on Acoustic–Vibration Fusion and Knowledge Transfer
by Fangyong Xue, Chang Liu, Feifei He and Zeping Bai
Sensors 2025, 25(16), 5190; https://doi.org/10.3390/s25165190 - 21 Aug 2025
Viewed by 551
Abstract
Although contact-based vibration signal methods for mechanical equipment fault diagnosis demonstrate superior performance, their practical deployment faces significant limitations. In contrast, acoustic signals offer notable deployment flexibility due to their non-contact nature. However, acoustic diagnostic methods are susceptible to environmental noise interference, and [...] Read more.
Although contact-based vibration signal methods for mechanical equipment fault diagnosis demonstrate superior performance, their practical deployment faces significant limitations. In contrast, acoustic signals offer notable deployment flexibility due to their non-contact nature. However, acoustic diagnostic methods are susceptible to environmental noise interference, and fault samples are typically scarce, leading to insufficient model generalization capability and robustness. To address this, this paper proposes an acoustic–vibration feature fusion strategy based on heterogeneous transfer learning, further integrated with a knowledge distillation framework. By doing so, it aims to achieve efficient transfer of vibration diagnostic knowledge to acoustic models. In the proposed approach, a teacher model learns diagnostic knowledge from highly reliable vibration signals and uses this to guide the training of a student model on acoustic signals. This process significantly enhances the diagnostic capability of the acoustic-based student model. Experimental studies conducted on a custom-built test rig and public datasets demonstrate that the proposed method exhibits excellent diagnostic accuracy and robustness under unseen working conditions. Full article
(This article belongs to the Section Fault Diagnosis & Sensors)
Show Figures

Figure 1

19 pages, 2983 KB  
Article
Detecting the Type and Severity of Mineral Nutrient Deficiency in Rice Plants Based on an Intelligent microRNA Biosensing Platform
by Zhongxu Li and Keyvan Asefpour Vakilian
Sensors 2025, 25(16), 5189; https://doi.org/10.3390/s25165189 - 21 Aug 2025
Viewed by 538
Abstract
The early determination of the type and severity of stresses caused by nutrient deficiency is necessary for taking timely measures and preventing a remarkable yield reduction. This study is an effort to investigate the performance of a machine learning-based model that identifies the [...] Read more.
The early determination of the type and severity of stresses caused by nutrient deficiency is necessary for taking timely measures and preventing a remarkable yield reduction. This study is an effort to investigate the performance of a machine learning-based model that identifies the type and severity of nitrogen, phosphorus, potassium, and sulfur in rice plants by using the plant microRNA data as model inputs. The concentration of 14 microRNA compounds in plants exposed to nutrient deficiency was measured using an electrochemical biosensor based on the peak currents produced during the probe–target microRNA hybridization. Subsequently, several machine learning models were utilized to predict the type and severity of stress. According to the results, the biosensor used in this work exerted promising analytical performance, including linear range (10−19 to 10−11 M), limit of detection (3 × 10−21 M), and reproducibility during microRNA measurement in total RNA extracted from rice plant samples. Among the microRNAs studied, miRNA167, miRNA162, miRNA169, and miRNA395 exerted the largest contribution in predicting the nutrient deficiency levels based on feature selection methods. Using these four microRNAs as model inputs, the random forest with hyperparameters optimized by the genetic algorithm was capable of detecting the type of nutrient deficiency with an average accuracy, precision, and recall of 0.86, 0.94, and 0.87, respectively, seven days after the application of the nutrient treatment. Within this period, the optimized machine was able to detect the level of deficiency with average MSE and R2 of 0.010 and 0.92, respectively. Combining the findings of this study and the results we reported earlier on determining the occurrence of salinity, drought, and heat in rice plants using microRNA biosensors can be useful to develop smart biosensing platforms for efficient plant health monitoring systems. Full article
Show Figures

Figure 1

37 pages, 1545 KB  
Article
BFL-SDWANTrust: Blockchain Federated-Learning-Enabled Trust Framework for Secure East–West Communication in Multi-Controller SD-WANs
by Muddassar Mushtaq and Kashif Kifayat
Sensors 2025, 25(16), 5188; https://doi.org/10.3390/s25165188 - 21 Aug 2025
Viewed by 662
Abstract
Software-Defined Wide-Area Networks (SD-WAN) efficiently manage and route traffic across multiple WAN connections, enhancing the reliability of modern enterprise networks. However, the performance of SD-WANs is largely affected due to malicious activities of unauthorized and faulty nodes. To solve these issues, many machine-learning-based [...] Read more.
Software-Defined Wide-Area Networks (SD-WAN) efficiently manage and route traffic across multiple WAN connections, enhancing the reliability of modern enterprise networks. However, the performance of SD-WANs is largely affected due to malicious activities of unauthorized and faulty nodes. To solve these issues, many machine-learning-based malicious-node-detection techniques have been proposed. However, these techniques are vulnerable to various issues such as low classification accuracy and privacy leakage of network entities. Furthermore, most operations of traditional SD-WANs are dependent on a third-party or a centralized party, which leads to issues such single point of failure, large computational overheads, and performance bottlenecks. To solve the aforementioned issues, we propose a Blockchain Federated-Learning-Enabled Trust Framework for Secure East–West Communication in Multi-Controller SD-WANs (BFL-SDWANTrust). The proposed model ensures local model learning at the edge nodes while utilizing the capabilities of federated learning. In the proposed model, we ensure distributed training without requiring central data aggregation, which preserves the privacy of network entities while simultaneously improving generalization across heterogeneous SD-WAN environments. We also propose a blockchain-based network that validates all network communication and malicious node-detection transactions without the involvement of any third party. We evaluate the performance of our proposed BFL-SDWANTrust on the InSDN dataset and compare its performance with various benchmark malicious-node-detection models. The simulation results show that BFL-SDWANTrust outperforms all benchmark models across various metrics and achieves the highest accuracy (98.8%), precision (98.0%), recall (97.0%), and F1-score (97.7%). Furthermore, our proposed model has the shortest training and testing times of 12 s and 3.1 s, respectively. Full article
(This article belongs to the Special Issue Sensors and IoT Technologies for the Smart Industry)
Show Figures

Figure 1

42 pages, 5531 KB  
Article
Preliminary Analysis and Proof-of-Concept Validation of a Neuronally Controlled Visual Assistive Device Integrating Computer Vision with EEG-Based Binary Control
by Preetam Kumar Khuntia, Prajwal Sanjay Bhide and Pudureddiyur Venkataraman Manivannan
Sensors 2025, 25(16), 5187; https://doi.org/10.3390/s25165187 - 21 Aug 2025
Viewed by 636
Abstract
Contemporary visual assistive devices often lack immersive user experience due to passive control systems. This study introduces a neuronally controlled visual assistive device (NCVAD) that aims to assist visually impaired users in performing reach tasks with active, intuitive control. The developed NCVAD integrates [...] Read more.
Contemporary visual assistive devices often lack immersive user experience due to passive control systems. This study introduces a neuronally controlled visual assistive device (NCVAD) that aims to assist visually impaired users in performing reach tasks with active, intuitive control. The developed NCVAD integrates computer vision, electroencephalogram (EEG) signal processing, and robotic manipulation to facilitate object detection, selection, and assistive guidance. The monocular vision-based subsystem implements the YOLOv8n algorithm to detect objects of daily use. Then, audio prompting conveys the detected objects’ information to the user, who selects their targeted object using a voluntary trigger decoded through real-time EEG classification. The target’s physical coordinates are extracted using ArUco markers, and a gradient descent-based path optimization algorithm (POA) guides a 3-DoF robotic arm to reach the target. The classification algorithm achieves over 85% precision and recall in decoding EEG data, even with coexisting physiological artifacts. Similarly, the POA achieves approximately 650 ms of actuation time with a 0.001 learning rate and 0.1 cm2 error threshold settings. In conclusion, the study also validates the preliminary analysis results on a working physical model and benchmarks the robotic arm’s performance against human users, establishing the proof-of-concept for future assistive technologies integrating EEG and computer vision paradigms. Full article
(This article belongs to the Section Intelligent Sensors)
Show Figures

Figure 1

11 pages, 299 KB  
Brief Report
Activity Type Effects Signal Quality in Electrocardiogram Devices
by Bryndan Lindsey, Samantha Snyder, Yuanyuan Zhou, Jae Kun Shim, Jin-Oh Hahn, William Evans and Joel Martin
Sensors 2025, 25(16), 5186; https://doi.org/10.3390/s25165186 - 20 Aug 2025
Viewed by 424
Abstract
Electrocardiogram (ECG) devices are commonly used to monitor heart rate (HR) and heart rate variability (HRV), but their signal quality under non-upright or torso-dominant activities may suffer due to motion artifact and interference from surrounding musculature. We compared ECG signal quality during treadmill [...] Read more.
Electrocardiogram (ECG) devices are commonly used to monitor heart rate (HR) and heart rate variability (HRV), but their signal quality under non-upright or torso-dominant activities may suffer due to motion artifact and interference from surrounding musculature. We compared ECG signal quality during treadmill walking, circuit training, and an obstacle course using three chest-worn commercial devices (Polar H10, Equivital EQ-02, and Zephyr BioHarness 3.0) and a multi-lead ECG monitor (BIOPAC). Signal quality was quantified using the beat signal quality index (SQI), and HR data were rejected if SQI fell below 0.7 or if values were physiologically implausible. Signal rejection rate was calculated as the proportion of low-quality observations across device and activity type. Significant effects of both device (p < 0.001) and activity (p < 0.001) were observed, with greater signal rejection during the obstacle course and circuit training compared to treadmill walking (p < 0.01). The Zephyr exhibited significantly higher rejection rates than the Polar (p = 0.018) and BIOPAC (p = 0.017), while the Polar showed lower average rejection rates across all activities. These findings underscore the importance of including dynamic, non-upright tasks in ECG validation protocols and suggest that certain commercial devices may be more robust under realistic conditions. Full article
(This article belongs to the Special Issue Biosignal Sensing Analysis (EEG, EMG, ECG, PPG) (2nd Edition))
Show Figures

Figure 1

Previous Issue
Back to TopTop