Search Results (1,536)

The Internet of Things (IoT) and Industrial Internet of Things (IIoT) have drastically transformed industries by enhancing efficiency and flexibility but have also introduced substantial cybersecurity risks. The rise of zero-day attacks, which exploit unknown vulnerabilities, poses significant threats to these interconnected systems. Traditional signature-based intrusion detection systems (IDSs) are insufficient for detecting such attacks due to their reliance on pre-defined attack signatures. This study investigates the effectiveness of Adaptive SAMKNN, an adaptive k-nearest neighbor with self-adjusting memory (SAM), in detecting and responding to various attack types in Internet of Things (IoT) environments. Through extensive testing, our proposed method demonstrates superior memory efficiency, with a memory footprint as low as 0.05 MB, while maintaining high accuracy and F1 scores across all datasets. The proposed method also recorded a detection rate of 1.00 across all simulated zero-day attacks. In scalability tests, the proposed technique sustains its performance even as data volume scales up to 500,000 samples, maintaining low CPU and memory consumption. However, while it excels under gradual, recurring, and incremental drift, its sensitivity to sudden drift highlights an area for further improvement. This study confirms the feasibility of Adaptive SAMKNN as a real-time, scalable, and memory-efficient solution for IoT and IIoT security, providing reliable anomaly detection without overwhelming computational resources. Our proposed method has the potential to significantly increase the security of IoT and IIoT environments by enabling the real-time, scalable, and efficient detection of sophisticated cyber threats, thereby safeguarding critical interconnected systems against emerging vulnerabilities. Full article

Seaweed is a valuable natural resource, but drift or beach-cast seaweed is considered a waste product. Although seaweed is traditionally used as an organic amendment, vermicomposting has the potential to transform the material into valuable organic fertilizer, thereby enhancing its microbial properties. This study aimed to investigate the dynamics of the taxonomic and functional bacterial communities in seaweed during the vermicomposting process by high-throughput sequencing of 16S rRNA gene amplicons. Vermicomposting changed the composition of the bacterial communities, as indicated by the low proportion of bacterial taxa common to the bacterial communities in the raw seaweed and vermicompost (21 to 56 ASVs from more than 900 ASVs per sample type). The observed increase in taxonomic diversity (32% mean increase across sampling times) also affected the functionality of the bacterial communities present in the vermicompost. The diverse bacterial community showed enriched functional pathways related to soil health and plant growth, including the synthesis of antibiotics, amino acids, and phytohormones, as well as the degradation of bisphenol. In conclusion, in terms of microbial load and diversity, vermicompost derived from seaweed is a more valuable organic fertiliser than seaweed itself. Full article

This study investigates the seismic performance of two-, four-, and six-story composite buildings equipped with viscous wall dampers, focusing on structures with concrete-filled steel tubular (CFST) columns and steel beams. Through nonlinear time history analyses using sequential ground motions, the research evaluates the effectiveness of viscous wall dampers in mitigating seismic demands. Results demonstrate significant reductions in both interstory drift ratios and peak floor accelerations across all building heights when dampers are installed. The study particularly highlights the dampers’ effectiveness in controlling drift demands in lower and middle floors while managing acceleration amplification at upper levels. The findings validate the integration of viscous wall dampers into mid-rise composite structures and underscore the importance of considering sequential ground motions in seismic performance evaluations. Full article

This study aimed to predict and fit the nonlinear dynamic grip force of the human upper limb using surface electromyographic (sEMG) signals. The research employed a time-series-based neural network, NARX, to establish a mapping relationship between the electromyographic signals of the forearm muscle groups and dynamic grip force. Three-channel electromyographic signal acquisition equipment and a grip force sensor were used to record muscle signals and grip force data of the subjects under specific dynamic force conditions. After preprocessing the data, including outlier removal, wavelet denoising, and baseline drift correction, the NARX model was used for fitting analysis. The model compares two different training strategies: regularized stochastic gradient descent (BRSGD) and conjugate gradient (CG). The results show that the CG greatly shortened the training time, and performance did not decline. NARX demonstrated good accuracy and stability in dynamic grip force prediction, with the model with 10 layers and 20 time delays performing the best. The results demonstrate that the proposed method has potential practical significance for force control applications in smart prosthetics and virtual reality. Full article

Classic enhanced self-correcting battery equivalent models require proper model parameters and initial conditions such as the initial state of charge for its unbiased functioning. Obtaining parameters is often conducted by optimization using evolutionary algorithms. Obtaining the initial state of charge is often conducted by measurements, which can be burdensome in practice. Incorrect initial conditions can introduce bias, leading to long-term drift and inaccurate state of charge readings. To address this, we propose two simple and efficient equivalent model frameworks that are optimized by a genetic algorithm and are able to determine the initial conditions autonomously. The first framework applies the feedback loop mechanism that gradually with time corrects the externally given initial condition that is originally a biased arbitrary value within a certain domain. The second framework applies the genetic algorithm to search for an unbiased estimate of the initial condition. Long-term experiments have demonstrated that these frameworks do not deviate from controlled benchmarks with known initial conditions. Additionally, our experiments have shown that all implemented models significantly outperformed the well-known ampere-hour coulomb counter integration method, which is prone to drift over time and the extended Kalman filter, that acted with bias. Full article

Transformer-based time series forecasting models use patch tokens for temporal patterns and variate tokens to learn covariates’ dependencies. While patch tokens inherently facilitate self-supervised learning, variate tokens are more suitable for linear forecasters as they help to mitigate distribution drift. However, the use of variate tokens prohibits masked model pretraining, as masking an entire series is absurd. To close this gap, we propose LSPatch-T (Long–Short Patch Transfer), a framework that transfers knowledge from short-length patch tokens into full-length variate tokens. A key implementation is that we selectively transfer a portion of the Transformer encoder to ensure the linear design of the downstream model. Additionally, we introduce a robust frequency loss to maintain consistency across different temporal ranges. The experimental results show that our approach outperforms Transformer-based baselines (Transformer, Informer, Crossformer, Autoformer, PatchTST, iTransformer) on three public datasets (ETT, Exchange, Weather), which is a promising step forward in generalizing time series forecasting models. Full article

Simultaneous localization and mapping (SLAM) techniques can be used to navigate the visually impaired, but the development of robust SLAM solutions for crowded spaces is limited by the lack of realistic datasets. To address this, we introduce InCrowd-VI, a novel visual–inertial dataset specifically designed for human navigation in indoor pedestrian-rich environments. Recorded using Meta Aria Project glasses, it captures realistic scenarios without environmental control. InCrowd-VI features 58 sequences totaling a 5 km trajectory length and 1.5 h of recording time, including RGB, stereo images, and IMU measurements. The dataset captures important challenges such as pedestrian occlusions, varying crowd densities, complex layouts, and lighting changes. Ground-truth trajectories, accurate to approximately 2 cm, are provided in the dataset, originating from the Meta Aria project machine perception SLAM service. In addition, a semi-dense 3D point cloud of scenes is provided for each sequence. The evaluation of state-of-the-art visual odometry (VO) and SLAM algorithms on InCrowd-VI revealed severe performance limitations in these realistic scenarios. Under challenging conditions, systems exceeded the required localization accuracy of 0.5 m and the 1% drift threshold, with classical methods showing drift up to 5–10%. While deep learning-based approaches maintained high pose estimation coverage (>90%), they failed to achieve real-time processing speeds necessary for walking pace navigation. These results demonstrate the need and value of a new dataset to advance SLAM research for visually impaired navigation in complex indoor environments. Full article

Robotic devices with integrated tactile sensors can accurately perceive the contact force, pressure, sliding, and other tactile information, and they have been widely used in various fields, including human–robot interaction, dexterous manipulation, and object recognition. To address the challenges associated with the initial value drift, and to improve the durability and accuracy of the tactile detection for a robotic dexterous hand, in this study, a flexible tactile sensor is designed with high repeatability by introducing a supporting layer for pre-separation. The proposed tactile sensor has a detection range of 0–5 N with a resolution of 0.2 N, and the repeatability error is as relatively small as 1.5%. In addition, the response time of the proposed tactile sensor under loading and unloading conditions are 80 ms and 160 ms, respectively. Moreover, a three-dimensional force decoupling detection method is developed by distributing tactile sensor units on a non-coplanar robotic fingertip. Finally, using a backpropagation neural network, the classification and recognition processes of nine types of objects with different shapes and categories are realized, achieving an accuracy higher than 95%. The results show that the proposed three-dimensional force tactile sensing system could be beneficial for the delicate manipulation and recognition for robotic dexterous hands. Full article

The general-purpose Compact Muon Solenoid (CMS) detector at the Large Hadron Collider (LHC) at CERN is used to study the production of new particles in proton–proton collisions at an LHC center of mass energy of 13.6 TeV. The detector includes a magnet based on a 6 m diameter superconducting solenoid coil operating at a current of 18.164 kA. This current creates a central magnetic flux density of 3.8 T that allows for the high-precision measurement of the momenta of the produced charged particles using tracking and muon subdetectors. The CMS magnet contains a 10,000 ton flux-return yoke of dodecagonal shape made from the assembly of construction steel blocks distributed in several layers. These steel blocks are magnetized with the solenoid returned magnetic flux and wrap the muons escaping the hadronic calorimeters of total absorption. To reconstruct the muon trajectories, and thus to measure the muon momenta, the drift tube and cathode strip chambers are located between the layers of the steel blocks. To describe the distribution of the magnetic flux in the magnet yoke layers, a three-dimensional computer model of the CMS magnet is used. To validate the calculations, special measurements are performed, with the flux loops wound in 22 cross-sections of the flux-return yoke blocks. The measured voltages induced in the flux loops during the CMS magnet ramp-ups and -downs, as well as during the superconducting coil fast discharges, are integrated over time to obtain the initial magnetic flux densities in the flux loop cross-sections. The measurements obtained during the seven standard ramp-downs of the magnet were analyzed in 2018. From that time, three fast discharges occurred during the standard ramp-downs of the magnet. This allows us to single out the contributions of the eddy currents, induced in steel, to the flux loop voltages registered during the fast discharges of the coil. Accounting for these contributions to the flux loop measurements during intentionally triggered fast discharges in 2006 allows us to perform the validation of the CMS magnet computer model with better precision. The technique for the flux loop measurements and the obtained results are presented and discussed. The method for measuring magnetic flux density in steel blocks described in this study is innovative. The experience of 3D modeling and measuring the magnetic field in steel blocks of the magnet yoke, as part of a muon detector system, has good prospects for use in the construction and operation of particle detectors for the Future Circular Electron–Positron Collider and the Circular Electron–Positron Collider. Full article

Effective monitoring of road conditions is crucial for ensuring safe and efficient transportation systems. By leveraging the power of crowd-sourced smartphone sensor data, road condition monitoring can be conducted in real-time, providing valuable insights for transportation planners, policymakers, and the general public. Previous studies have primarily focused on the use of pre-trained machine learning models and threshold-based methods for anomaly classification, which may not be suitable for real-world scenarios that require incremental detection and classification. As a result, there is a need for novel approaches that can adapt to changing data environments and perform effective classification without relying on pre-existing training data. This study introduces a novel, real-time road condition monitoring technique harnessing smartphone sensor data, addressing the limitations of pre-trained models that lack adaptability in dynamic environments. A hybrid anomaly detection method, combining unsupervised and supervised learning, is proposed to effectively manage concept drift, demonstrating a significant improvement in accuracy and robustness with a 96% success rate. The findings underscore the potential of incremental learning to enhance model responsiveness and efficiency in distinguishing various road anomalies, offering a promising direction for future transportation safety and resource optimization strategies. Full article

With the increasing demand for oil and gas, the depth of some vertical gas wells can reach 6000 m. At this time, the downhole fluid is in a state of high temperature and pressure, and interpretation of the production logging output profile faces the problem of inaccurate production calculations and difficulty judging the water-producing layer. The drift-flux model is usually used to calculate the gas–water two-phase flow. The drift-flux model is widely used to describe the two-phase flow in pipelines and wells because of its accuracy and simplicity. The constitutive correlations used in drift-flux models, which specify the relative motion between phases, have been extensively studied. However, most of the correlations are only extended by laboratory data of small pipe diameters at standard temperature and pressure and do not apply to high-temperature and high-pressure large-diameter gas wells. Therefore, we improved the distribution coefficient and drift velocity of drift-flux correlations in this study for high-temperature and high-pressure gas wells with large pipe diameters. Therefore, this study improved the distribution coefficient and drift velocity of the drift-flux correlations for high-temperature and high-pressure gas wells with large pipe diameters. In practical application, the coincidence rates of gas production and water production calculated by the new drift-flux model were 12.68% and 19.39%, respectively. For high-temperature and high-pressure deep wells, the measurement errors of production logging instruments are significant, and surface laboratory pipelines are challenging to configure and equip with actual high-temperature and high-pressure conditions. Therefore, this study used the method of numerical simulation to study the flow characteristics of the two phases of high-temperature and high-pressure gas and water to provide a basis for identifying the water layer. Combined with the proposed drift-flux correlations and the new method of determining the water-producing layer, a new method of production profile interpretation of high-temperature and high-pressure gas wells is obtained. Full article

Seismic fragility curves, derived from ground motion data, are essential tools for predicting and assessing potential damage to structures during earthquakes. Seismic fragility curves are vital for assessing the structural behavior of buildings and establishing disaster response criteria when an earthquake occurs. We performed an incremental dynamic analysis based on 400 ground motion data. We sampled various sets of ground motions (10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 150, 200, 250, 300, and 350) and derived seismic fragility curves for three performance criteria, based on inter-story drift, by conducting 100,000 simulations for two steel frame structures each (6-story and 13-story). Fewer ground motions increase the uncertainty of the seismic fragility curve, distorting the results. Conversely, increasing the number of ground motions improves the reliability of the input variables and enhances the consistency of the results. The median and the logarithmic standard deviation for both structures converged toward the reference values when 30 or more ground motions were used. Similar results were observed when ≥50 ground motions were used. Specifically, more ground motions corresponded with a lower uncertainty in deriving the input variables for the seismic fragility curve, improving the reliability of the results. In conclusion, the number of ground motions used is directly related to the computational time for numerical analysis when deriving seismic fragility curves. Therefore, considering an appropriate number of ground motions is crucial to enhancing the reliability of the input variables used in evaluating the structural performance. Full article

The use of ion-selective field-effect transistors (ISFETs) facilitates real-time nutrient analysis in agricultural applications, including soil analysis and hydroponics. The rapid digital availability of analysis results allows for the implementation of ion-specific fertilisation control. The success, accuracy, and robustness of measurements using ISFET technology strongly depend on the handling of the process. This article presents a detailed overview of the sub-process steps required for the implementation of a stable automated application-specific ISFET-based measurement. This article provides experience-based recommendations for handling the conditioning, full calibration, and single-point calibration of the ISFET sensors. The hypotheses were empirically tested under authentic conditions and subsequently integrated into an overall process optimisation strategy. A comprehensive investigation has been conducted with the objective of gaining a deeper understanding of the ISFET baseline drift and implementing corrective measures. The results show that the baseline drift can be quantified and taken into account in the evaluation of the ISFET measurements. The efficacy of these measures was validated using standard laboratory analyses. Full article

The introduction of Self-X capabilities into industrial control offers a tremendous potential in the development of resilient, adaptive production systems that enable circular economy. The Self-X capabilities, powered by Artificial Intelligence (AI), can monitor the production performance and enable timely reactions to problems or suboptimal operation. This paper presents a concept and prototype for Self-X AI in the process industry, particularly electric steelmaking with the EAF (Electric Arc Furnace). Due to complexity, EAF operation should be optimized with computational models, but these suffer from the fluctuating composition of the input materials, i.e., steel scrap. The fluctuation can be encountered with the Self-X method that monitors the performance, detecting anomalies and suggesting the re-training and re-initialization of models. These suggestions support the Human-in-the-Loop (HITL) in managing the AI models and in operating the production processes. The included Self-X capabilities are self-detection, self-evaluation, and self-repair. The prototype proves the concept, showing how the optimizing AI pipeline receives alarms from the external AI services if the performance degrades. The results of this work are encouraging and can be generalized, especially to processes that encounter drift related to the conditions, such as input materials for circular economy. Full article

The demand for non-invasive, real-time health monitoring has driven advancements in wearable sensors for tracking biomarkers in sweat. Ammonium ions (NH₄⁺) in sweat serve as indicators of metabolic function, muscle fatigue, and kidney health. Although current ion-selective all-solid-state printed sensors based on nanocomposites typically exhibit good sensitivity (~50 mV/log [NH₄⁺]), low detection limits (LOD ranging from 10⁻⁶ to 10⁻⁷ M), and wide linearity ranges (from 10⁻⁵ to 10⁻¹ M), few have reported the stability test results necessary for their integration into commercial products for future practical applications. This study presents a highly stable, wearable electrochemical sensor based on a composite of metal–organic frameworks (MOFs) and reduced graphene oxide (rGO) for monitoring NH₄⁺ in sweat. The synergistic properties of Ni-based MOFs and rGO enhance the sensor’s electrochemical performance by improving charge transfer rates and expanding the electroactive surface area. The MOF/rGO sensor demonstrates high sensitivity, with a Nernstian response of 59.2 ± 1.5 mV/log [NH₄⁺], an LOD of 10^−6.37 M, and a linearity range of 10⁻⁶ to 10⁻¹ M. Additionally, the hydrophobic nature of the MOF/rGO composite prevents water layer formation at the sensing interface, thereby enhancing long-term stability, while its high double-layer capacitance minimizes potential drift (7.2 µV/s (i = ±1 nA)) in short-term measurements. Extensive testing verified the sensor’s exceptional stability, maintaining consistent performance and stable responses across varying NH₄⁺ concentrations over 7 days under ambient conditions. On-body tests further confirmed the sensor’s suitability for the continuous monitoring of NH₄⁺ levels during physical activities. Further investigations are required to fully elucidate the impact of interference from other sweat components (such as K⁺, Na⁺, Ca²⁺, etc.) and the influence of environmental factors (including the subject’s physical activity, posture, etc.). With a clearer understanding of these factors, the sensor has the potential to emerge as a promising tool for wearable health monitoring applications. Full article