Continuous Data Processing Research Articles

Application of Machine Learning Algorithms in Real-Time Monitoring of Conveyor Belt Damage

This paper is devoted to the real-time monitoring of close transportation devices, namely, belt conveyors. It presents a novel measurement system based on the linear strain gauges placed on the tail pulley surface. These gauges enable the monitoring and continuous collection and processing of data related to the process. An initial assessment of the machine learning application to the load identification was made. Among the tested algorithms that utilized machine learning, some exhibited a classification accuracy as high as 100% when identifying the load placed on the moving belt. Similarly, identification of the preset damage was possible using machine learning algorithms, demonstrating the feasibility of the system for fault diagnosis and predictive maintenance.

Numerical study of the influence of temperature data error on the determination of frictional heat generation

Subject of research: determination of the friction torque and functions of the specific intensity of heat generation in system of sliding bearings. The purpose of the study: To develop a method for determining the friction torque in sliding bearings using temperature data. Based on computational experiments, study the influence of errors in temperature data on the solution of the inverse problem. Methods and object of research: A system of sliding bearings made of a polymer composite material on which a rotating shaft rests is considered. A mathematical model of the thermal process in the considered system of friction units is presented, taking into account the spatial distribution of temperature and its change over time. The temperature is measured at several points in each bearing. Frictional heat generation is determined by solving the inverse heat conduction problem from the condition that the measured and calculated temperatures are close. To ensure continuous data processing and determine the friction torque during long-term tests, the inverse problem of determining the friction torque from temperature data is solved at successive short time intervals. Then the resulting solutions are «glued together». The main results of the study: With such a restoration, the discrepancy between the specified and restored functions of the specific intensity of heat generation is 10–15 % with an error level in temperature data of 10 %. The developed algorithm for solving the inverse problem of thermal conductivity can be used to determine the friction torques in real systems of self-lubricating sliding bearings.

Machine Vision AI in Railroad Safety: Advanced Inspection Techniques

This article explores the transformative impact of machine vision AI in railroad safety, focusing on advanced inspection techniques. It examines the integration of technologies such as line scan cameras, LiDAR, and IoT sensors in enabling comprehensive 360-degree inspections of railway infrastructure and rolling stock. The article discusses key innovations in continuous monitoring, data processing, and AI-driven analysis, highlighting their potential to significantly enhance defect detection accuracy, reduce maintenance costs, and prevent accidents. Real-world applications in structural integrity assessment and overhead line inspection are presented, along with an analysis of current challenges and future developments in the field. The article underscores the paradigm shift from reactive to proactive maintenance strategies, promising a safer and more efficient future for rail transport.

Organic Polymer-Based Photodiodes for Optoelectronic Reservoir Computing with Time-Based Coding.

The integration of optoelectronic devices with reservoir computing offers a novel and effective approach to in-sensor computing. This work presents a hybrid digital-physical solution that leverages the high-performance poly[(bithiophene)-alternate-(2,5-di(2-octyldodecyl)-3,6-di(thienyl)-pyrrolyl pyrrolidone)] (DPPT-TT) organic polymer-based photodiodes for the hardware implementation of reservoir computing system. The photodiodes demonstrate nonlinear photoelectric responses, fading memory, and cyclical stability, in relation to the temporal information on light stimuli. These attributes enable effective mapping, historical context sensitivity, and consistent performance, with time-encoded inputs, which are particularly essential for accurate and continuous processing of time series data. The optoelectronic reservoir computing system with pulse width modulation (PWM) coding demonstrates impressive performance in the prediction of chaotic sequences, achieving a normalized root-mean-square error as low as 0.095 with optimized parameters. The DPPT-TT-based photodiodes and time-based coding offer a hardware-efficient solution for reservoir computing, significantly advancing Internet of Things applications.

Analisis Deformasi di Lampung dan Selat Sunda berdasarkan Data GNSS tahun 2018 hingga 2021

The Lampung Province and Sunda Strait have a seismic gap zone with the potential for major earthquakes in the future. This study analyzes the deformation occurring in this region using continuous Global Navigation Satellite System (GNSS) station data from Indonesia Continuously Operating Reference Station (InaCORS) and Sumatran GPS Array (SuGAr) from 2018 to 2021.5. The GNSS data was processed using the Bernese 5.2 scientific software, applying least squares for velocity changes and statistical tests to analyze significance. The data processing was carried out in two schemes: the first scheme covering 2018-2020, and the second covering 2019-2021. The results of the deformation analysis from 2018 to 2021, using two continuous GNSS data processing schemes, showed velocity changes relative to the Sundaland Plate ranging from ~2 mm/year to ~20 mm/year. In the eastern region of the Sumatra fault, the velocity changes were smaller, around ~5 mm/year, due to the minimal influence of tectonic activity. However, in the Sunda Strait region, the deformation was influenced by volcanic activity. The deformation occurring in Lampung Province and the Sunda Strait, based on GNSS velocity changes, significantly contributes to tectonic and volcanic activities.

Automated identification and long-term tracking of modal parameters for a super high-rise building

Structural modal parameter identification plays an important role in wind resistance design, damage identification, and health monitoring. At present, the modal identification process still suffers from low automation and low computational efficiency, which are not conducive to the continuous processing of monitoring data. To address this, an automated modal parameter identification approach is proposed by introducing the fast density and grid based (DGB) clustering, and applied to the Shanghai Tower, which is the tallest building in China. Moreover, in order to determine the optimal model order of the stochastic subspace identification (SSI), the trend of the logarithmic reduction rate for singular values is extracted by empirical mode decomposition (EMD). The numerical simulation of a cantilever beam and the field measurement of Shanghai Tower under the influence of Typhoon In-Fa are conducted to validate the effectiveness of the proposed approach. Results demonstrate that the approach can obtain modal parameters without human intervention and has the advantages of simple process, fast computation and high precision. Furthermore, based on the data collected from the structural health monitoring system of the Shanghai Tower over a period of 5.5 years, the approach is applied to analyze the correlation between natural frequencies and ambient temperature, as well as the long-term evolution law of modes, discovering that multi-order frequencies show a decreasing trend over time. The finite element analysis suggests that this fact might be attributed to the increase in structural mass caused by the rising usage rate and the degradation in material properties.

Pemanfaatan Sistem Informasi Manajemen dalam Pengelolaan Data Peserta Didik

This study aims to analyze the utilization of management information systems in managing student data at MA Al Hidayah Bakke using a qualitative approach. The primary data source was the madrasah operators. Data collection involved observation, interviews, and documentation. Data analysis techniques included data reduction, presentation, and drawing conclusions. The research findings indicate that the utilization of the management information system at MA Al Hidayah Bakke has integrated components such as hardware, software, and human resources to enhance the efficiency of madrasah data management. Support from thorough planning, facilities, and strict control ensures the quality and accuracy of the data. The data management process encompasses student data collection, continuous data processing in EMIS and SIMPATIKA systems, and data presentation through formal reports. By utilizing laptops to access EMIS and SIMPATIKA, the MIS significantly contributes to producing rapid and accurate reports, thereby improving the quality of educational services and the effectiveness of monitoring student development.

Brain-Controlled Computer Tasks for Paralyzed Persons: Framework Overview

To bridge the digital barrier between those with disabilities and those without impairments, several applications can be operated by brain signals. It highlights the importance of employing brain-computer interfaces (BCI) and mentions a specific system called the "Brain-Controlled Computer System" that can control a computer cursor and perform mouse actions using brain signals.. The system can control the cursor and provide mouse left/right click with brain signals using a low-cost Brainlink Lite device. Based on this system, we employed machine learning and deep learning methodologies to develop a system that classifies brain signals for cursor movement. The methodology involved gathering user input through surveys, interviews, and comparison studies, as well as utilizing pre-developed datasets collected using a mind sensor. We initially trained a model on a previous dataset, achieving approximately 70% accuracy. However, when applied to the new dataset, the accuracy dropped to 49%. To improve the results, we experimented with more complex models, including additional layers and a dropout layer to prevent overfitting. Despite these enhancements, the accuracy only reached a maximum of 52%. The research team collected a dataset tailored to their specific application. They made a groundbreaking discovery by applying a convolutional neural network (CNN), typically used for image analysis, to process continuous brainwave data. Despite this non-standard use, their hypothesis proved correct, with the CNN model achieving an impressive 80% accuracy in classifying brain signals for controlling a cursor.

Enhancing Forest Fire Risk Assessment: An Ontology-Based Approach with Improved Continuous Apriori Algorithm

Forest fires are sudden and difficult to extinguish, so early risk assessment is crucial. However, there are currently a lack of suitable knowledge-mining algorithms for forest fire risk assessment. This article proposes an improved continuous Apriori algorithm to mining forest fire rules by introducing prior knowledge to classify input data and enhance its ability to process continuous data. Meanwhile, it constructs an ontology to provide a standardized expression platform for forest fire risk assessment. The improved continuous Apriori algorithm cooperates with ontology and applies the mining rules to the forest fire risk assessment results. The proposed method is validated using the forest fire data from the Bejaia region in Algeria. The results show that the improved continuous Apriori algorithm is superior to the raw Apriori algorithm and can mine the rules ignored by the raw Apriori algorithm. Compared to the raw Apriori algorithm, the number of generated rules increased by 191.67%. The method presented here can be used to enhance forest fire risk assessments and contribute to the generation and sharing of forest-fire-related knowledge, thereby alleviating the problem of insufficient knowledge in forest fire risk assessment.

Innova4Health: an integrated approach for prevention of recurrence and personalized treatment of Major Depressive Disorder.

Major Depressive Disorder (MDD) is a prevalent mental health condition characterized by persistent low mood, cognitive and physical symptoms, anhedonia (loss of interest in activities), and suicidal ideation. The World Health Organization (WHO) predicts depression will become the leading cause of disability by 2030. While biological markers remain essential for understanding MDD's pathophysiology, recent advancements in social signal processing and environmental monitoring hold promise. Wearable technologies, including smartwatches and air purifiers with environmental sensors, can generate valuable digital biomarkers for depression assessment in real-world settings. Integrating these with existing physical, psychopathological, and other indices (autoimmune, inflammatory, neuroradiological) has the potential to improve MDD recurrence prevention strategies. This prospective, randomized, interventional, and non-pharmacological integrated study aims to evaluate digital and environmental biomarkers in adolescents and young adults diagnosed with MDD who are currently taking medication. The study implements a sensor-integrated platform built around an open-source "Pothos" air purifier system. This platform is designed for scalability and integration with third-party devices. It accomplishes this through software interfaces, a dedicated app, sensor signal pre-processing, and an embedded deep learning AI system. The study will enroll two experimental groups (10 adolescents and 30 young adults each). Within each group, participants will be randomly allocated to Group A or Group B. Only Group B will receive the technological equipment (Pothos system and smartwatch) for collecting digital biomarkers. Blood and saliva samples will be collected at baseline (T0) and endpoint (T1) to assess inflammatory markers and cortisol levels. Following initial age-based stratification, the sample will undergo detailed classification at the 6-month follow-up based on remission status. Digital and environmental biomarker data will be analyzed to explore intricate relationships between these markers, depression symptoms, disease progression, and early signs of illness. This study seeks to validate an AI tool for enhancing early MDD clinical management, implement an AI solution for continuous data processing, and establish an AI infrastructure for managing healthcare Big Data. Integrating innovative psychophysical assessment tools into clinical practice holds significant promise for improving diagnostic accuracy and developing more specific digital devices for comprehensive mental health evaluation.

Internet Users' Willingness to Disclose Biometric Data for Continuous Online Account Protection: An Empirical Investigation

Continuous authentication has emerged as a promising approach to increase user account security for online services. Unlike traditional authentication methods, continuous authentication provides ongoing security throughout the session, protecting against session takeover attacks due to illegitimate access. The effectiveness of continuous authentication systems relies on the continuous processing of users' sensitive biometric data. To balance security and privacy trade-offs, it's crucial to understand when users are willing to disclose biometric data for enhanced account security, addressing inevitable privacy concerns and user acceptance. To address this knowledge gap, we conducted an online study with 830 participants from the U.S., aiming to investigate user perceptions towards continuous authentication across different classes of online services. Our analysis identified four groups of biometric traits that directly reflect users' willingness to disclose them. Our findings demonstrate that willingness to disclose is influenced by both the specific biometric traits and the type of online service involved. User perceptions are strongly shaped by factors such as response efficacy, perceived privacy risks associated with the biometric traits, and concerns about the service providers' handling of such data. Our results emphasize the inadequacy of one-size-fits-all solutions and provide valuable insights for the design and implementation of continuous authentication systems.

Certain Investigations of prediction on Stock trend using various Optimization Techniques

A stock price represents a company’s value at any given point, trends of the same will be very volatile because of different trading activities, supply and demand of stocks, and companies’ financial outcomes. Predicting the correlation between price, time, and various other variables in any stock trend is an essential need for portfolio optimization. The model of LSTM(Long Short Term Memory) recurrent neural networks (RNN) is the optimal prediction method, with LSTM used for understanding temporal dependencies, which is well known for processing and understanding continuous data points, The above model gives structural integrity to most of the time-series data analysis. The stock market produces a vast amount of data, there will be fluctuation of prices every second, so training Neural Networks for an enormous amount of data takes extensive time, We are performing certain investigations on boosting the accuracy and reducing the time taken to train by further enhancing the above-given model, with modified versions of Adam, RMSProp, and AdaGrad optimization methods.

Metasurface enabled broadband all optical edge detection in visible frequencies

Image processing is of fundamental importance for numerous modern technologies. In recent years, due to increasing demand for real-time and continuous data processing, metamaterial and metasurface based all-optical computation techniques emerged as a promising alternative to digital computation. Most of the pioneer research focused on all-optical edge detection as a fundamental step of image processing. Metasurfaces have been shown to enable real time edge detection with low to no power consumption. However, the previous demonstrations were subjected to the several limitations such as need for oblique-incidence, polarization dependence, need for additional polarizers, narrow operation bandwidth, being limited with processing in 1D, operation with coherent light only, and requiring digital post-processing. Here, we propose and experimentally demonstrate 2D isotropic, polarization-independent, broadband edge detection with high transmission efficiency under both coherent and incoherent illumination along the visible frequency range using a metasurface based on Fourier optics principles.

FIDEL: Fog integrated federated learning framework to train neural networks

AbstractTechnological advancement in the digital era has continued to produce voluminous amounts of data through various devices. Even though data is produced distributively, it needs to be accumulated centrally for processing, analysis, and knowledge extraction that faces several challenges such as bandwidth, latency, congestion, privacy, and security. Fog computing paradigm addresses some of these issues, and can be used as a distributed data processing unit. Federated learning trains a shared model over distributed nodes. However, a fog node can not process continuously growing data due to computational limitations. In this paper, we propose FIDEL: a fog integrated federated learning framework for neural network training using resource‐constrained devices. The federation of resource‐constrained Internet of Things (IoT) devices creates a shared global model trained on local data, which is generalized on the unseen dataset for prediction/inferences. We have also designed an online training scheme to process continuous data with limited compute resources. The FIDEL supports both synchronous and asynchronous federate learning that empowers resource‐constrained devices to train machine learning models. To test the learning capabilities of the FIDEL, we have trained three neural networks (i) Shallow network; (ii) Deep Network; (iii) Convolutional Neural Network (CNN) models for human position detection in industrial IoT setup on rapidly changing datasets. The experimental results show that the framework can learn input–output relationships with significantly high accuracy. The overall system efficiency of the framework is reasonable in terms of latency and memory usage for resource‐constrained devices.

Hardware Software Co-Design for Impact Localization using Hybrid Laminates

Impact detection using piezoelectric sensors is an actual and widespread research field. The current work provides an approach for a real-time realization of an impact detection system using deep learning methods. For realization a hardware software co-design approach is used utilizing hardware acceleration by a continuous pipelining FPGA structure. The concept describes the hardware software partitioning of the underlying functions and the methodology for ensuring continuous data processing and the associated real-time capability. The behavior of the hardware is realized with the help of a finite state machine and thus the correctness of the data is ensured and the impact identification is realized. The results show the real-time capability as well as a reasonable resource utilization of the FPGA design.

The February 2018 seismic swarm in São Miguel, Azores

The island of São Miguel is among the most seismically active areas of the Azores archipelago. This work focuses on the most significant recent swarm, which occurred on February 2018. We set up an automated procedure to process continuous full seismic waveform data from local stations to generate high-quality earthquake information on the volcano unrest episode. First, we applied an automated detector software, next we located the detected events and then classified the earthquakes based on their waveform similarity, identifying three families of seismic events. We then extended the catalog by template matching. Finally, we computed moment tensors to investigate the source mechanisms of the largest earthquakes. Our results image the ∼2-week swarm evolution. The activity started with a precursory phase with low rate and low magnitude (ML < 2.0) seismicity and the activation of a deeper structure (∼10-15 km). After ∼1 week, a new earthquake family emerged at shallower depths (∼8–12 km) reaching magnitudes up to ML 3.4. Finally, a third slightly shallower family was activated. Moment tensors show mostly normal faulting mechanisms, striking ∼NW-SE, compatible with the orientation of the regional stress field. A surface deformation transient was recorded by geodetic stations, starting with the swarm, and continuing over the following ∼17 months, corresponding to either inflation or extension around the swarm region. The prolonged surface deformation implies a process that was initiated during the swarm and subsequently accommodated mostly aseismically. We interpret the seismicity observed at the early stage of deformation as indicating episodic fluid injection through the crust, related to the local hydrothermal or magmatic systems. We conclude that the Fogo-Congro region continues to be seismo-volcanically active, with both seismic and aseismic deformation observed and requiring close multidisciplinary monitoring. The proposed methology based on the automated analysis of continuous waveform data provides high-quality imaging of the spatio-temporal evolution of seismicity, which can be used elsewhere in the operational monitoring of seismo-volcanic crises to gain insight into the ongoing deformation processes, improve hazard assessment and help in the development of effective mitigation strategies.

Dynamic Network-Assisted D2D-Aided Coded Distributed Learning

Today, numerous machine learning (ML) applications offer continuous data processing and real-time data analytics at the edge of wireless networks. Distributed real-time ML solutions are highly susceptible to the so-called straggler effect caused by resource heterogeneity, which can be mitigated by various computation offloading mechanisms that severely impact communication efficiency, especially in large-scale scenarios. To reduce the communication overhead, we leverage device-to-device (D2D) connectivity, which enhances spectrum utilization and allows for efficient data exchange between proximate devices. In particular, we design a novel D2D-aided coded distributed learning method named D2D-CDL for efficient load balancing across devices. The proposed solution captures system dynamics, including <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">data</i> (time-varying learning model, irregular intensity of data arrivals), <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">device</i> (diverse computational resources and volume of training data), and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">deployment</i> (different locations and D2D graph connectivity). To decrease the number of communication rounds, we derive an optimal compression rate, which minimizes the processing time. The resulting optimization problem provides suboptimal compression parameters that improve the total training time. Our proposed method is particularly beneficial for real-time collaborative applications, where users continuously generate training data thus yielding a model drift.

On the building of efficient self-adaptable health data science services by using dynamic patterns

Health data science systems are becoming key for supporting healthcare decision-making processes. However, these systems should achieve continuous data processing and adapt their behavior to changes arising in real scenarios. Nevertheless, building this type of self-adaptable systems is not trivial, as it requires integrating data analytics and artificial intelligence with cloud computing tools and security and fault tolerance applications. This produces delays in the processing of data, affecting the information delivery in decision-making processes. In this paper, we present the design and implementation of a method to build self-adaptable data science services by using dynamic patterns. This method includes two models: a construction model and a coupling model. The construction model is based on dynamic parallel patterns to create, at design time, in-memory processing structures including as many applications as needed to meet the non-functional requirements (NFRs), such as security and fault tolerance, established by healthcare organizations without affecting the efficiency of a data science system. In turn, the coupling model converts the in-memory processing structures into software blocks that are coupled with the I/O interfaces of data science systems to support the automatic and transparent continuous management of data for facing changes in the incoming workload during execution time. A prototype was implemented to create self-adaptable health data science systems including in-memory processing structures for managing spirometry studies, tomography images, and electrocardiograms. The performance evaluation showed that the dynamic patterns significantly reduced the response time required to prepare contents with NFR characteristics in comparison with solutions from the state-of-the-art such as Nextflow and Makeflow.

A Mitigation Strategy for the Prediction Inconsistency of Neural Phase Pickers

AbstractNeural phase pickers—neural networks designed and trained to pick seismic phase arrivals—have proven to be a powerful tool for developing earthquake catalogs. However, these pickers suffer from prediction inconsistency in which the results they produce change, sometimes substantially, even under a small perturbation to the input waveform. This problem has not been addressed by the developers and users of these pickers. In this study, we show how prediction inconsistency can negatively affect the completeness of earthquake catalogs developed using neural phase pickers. We show that simply using a small step size for the sliding window when processing continuous waveform data and aggregating the results significantly mitigates this problem. We also highlight the importance of training datasets for increasing the consistency and other performance metrics.

Near-optimal adaptive predictive control model study for roller shades in office spaces

As improved building energy efficiency becomes mandatory, dynamic shading systems are expected to reduce building energy demand because their control mechanisms are increasingly self-adaptable to changes in the environment. Various shading control strategies had been proposed and compared in existing literature, but their energy performance can be further improved with a more comprehensive consideration of environmental changes and indoor demand. This study develops and validates an adaptive predictive control (APC) model for roller shades that combines simulation-based optimization of shading control with machine learning. The model predicts the appropriate position of roller shades based on multiple environmental parameters, minimizing the building energy demand. EnergyPlus and Python provide continuous energy simulation and data processing. Using pre-simulated datasets, machine learning algorithms develop APC models. The study scenario is a typical office space in Guangzhou, China. Quantitative analysis of the simulation results elucidates the effectiveness of the method. The performance of the model was evaluated using historical meteorological data from 2020. The proposed control model was able to achieve nearly optimal energy efficiency, reducing energy demand by 38.3% while reaching annual Useful Daylight Illuminances of 72.9%.