Data Science in Engineering Vol. 10
Proceedings of the 42nd IMAC, A Conference and Exposition on Structural Dynamics 2024
- 2025
- Book
- Editors
- Thomas Matarazzo
- François Hemez
- Eleonora Maria Tronci
- Austin Downey
- Publisher
- Springer Nature Switzerland
About this book
Data Science in Engineering, Volume 10: Proceedings of the 42nd IMAC, A Conference and Exposition on Structural Dynamics, 2024, the tenth volume of ten from the Conference brings together contributions to this important area of research and engineering. The collection presents early findings and case studies on fundamental and applied aspects of Data Science in Engineering, including papers on:
Novel Data-driven Analysis Methods Deep Learning Gaussian Process Analysis Real-time Video-based Analysis Applications to Nonlinear Dynamics and Damage Detection Data-driven System Prognostics
Advertisement
Table of Contents
-
Frontmatter
-
Physics-Informed Machine Learning Part I: Different Strategies to Incorporate Physics into Engineering Problems
Eleonora Maria Tronci, Austin R. J. Downey, Azin Mehrjoo, Puja Chowdhury, Daniel CobleThe chapter delves into the integration of physics into machine learning models for engineering problems, focusing on the advantages and challenges of physics-informed machine learning (PIML). It categorizes PIML strategies based on data availability and the incorporation of physical principles, highlighting seven key approaches: data augmentation, delta learning-physics corrector, physics-constrained, physics-guided, transfer learning, delta learning-unknown physics, and physics-encoded. Each strategy is explained with practical examples, illustrating how PIML can improve model accuracy, generalization, and interpretability. The chapter also discusses open research questions and future directions in PIML, emphasizing the need for uncertainty quantification, robust model design, and the expansion of PIML methods to new applications and machine learning techniques. This detailed exploration of PIML strategies and their real-world applications makes the chapter an essential read for professionals seeking to enhance their understanding and use of PIML in engineering problems.AI Generated
This summary of the content was generated with the help of AI.
AbstractPhysics-informed machine learning (PIML) is a methodology that combines principles from physics with machine learning (ML) techniques to enhance the accuracy and interpretability of predictive models. By incorporating physical laws and constraints into the learning process, physics-informed machine learning enables more robust predictions and reduces the need for large amounts of training data. PIML has a wide range of applications in science and engineering, such as modeling physical systems, solving partial differential equations, and performing inverse analysis and optimization.In part I of this two-part series, the authors will provide attendees with an overview of the main concepts, methods, applications, and challenges of PIML. According to the way that a first-principle model is integrated with a data-driven ML model, it is possible to classify physics-informed strategies. In this overview, seven strategies will be covered: physics-constrained ML; physics-guided ML; physics-encoded ML; data-augmentation via physics principles; transfer learning from physics-based synthetic data to experimental data; delta-learning physics correction to improve physics generalization and delta-learning unknown physics to represent unmodeled physical phenomena. The benefits of these approaches including better generalization, explainability, and efficiency of the ML models will be addressed. This work will present related challenges and limitations of each approach. Finally, the authors will discuss some open research questions and future directions for PIML. By the end of this tutorial, the participants will have a comprehensive understanding of the principles and potential of PIML, as well as the ability to critically evaluate PIML models. -
Statistical Evaluation of Machine Learning for Vibration Data
Samuel Myren, Nidhi Parikh, Garrison Flynn, Dave Higdon, Emily CasletonThe chapter 'Statistical Evaluation of Machine Learning for Vibration Data' delves into the challenges and methods for evaluating machine learning models, particularly foundation models, in seismic data. It introduces the concept of phase picking in seismology and discusses the limitations of traditional machine learning approaches. The authors propose a novel evaluation framework for foundation models and demonstrate its application using deep learning models and benchmark seismic datasets. The chapter also explores the impact of data quality and sampling rates on model performance and provides insights into the interpretability and generalizability of these models. The research highlights the potential of foundation models in signal-related fields and suggests avenues for future work in this area.AI Generated
This summary of the content was generated with the help of AI.
AbstractResearch in machine learning (ML) for structural health monitoring (SHM) has increased in recent years due to ML’s potential to extract information from large quantities of data collected under varying conditions to make predictions, such as detection, identification, and characterization. Current ML implementation consists of training a model for narrow, task-specific needs, such as training a neural network to detect the presence of structural faults. However, advancements in ML and artificial intelligence (AI) have contributed foundation models—large models trained on broad data in a self-supervised fashion that can be quickly fine-tuned to accomplish specific downstream tasks different from the task it was trained on. For example, a foundation model that is trained for damage detection may be prompted to perform identification and characterization tasks given future data. Because the training and tasking of foundation models are inherently different from current ML implementation, metrics that capture performance while also considering these differences are needed for fair evaluation and comparison between approaches. This research focuses on the testing, evaluation, and benchmarking for foundation models trained on vibration data from ground sensors. We will present metrics that incorporate a broad series of tasks to quantify the performance of foundation models against other traditional ML models. Since the ground vibration data is employed to train and test foundation models for seismic events, the tasks of these models are well aligned with SHM: detecting, identifying, and characterizing acceleration signals or images. The comprehensive evaluation approach and metrics we present will be a step toward holistically quantifying advanced ML/AI success as it inevitably permeates the field of SHM. -
Quantifying the Value of Information Transfer in Population-Based SHM
A. J. Hughes, J. Poole, N. Dervilis, P. Gardner, K. WordenThis chapter delves into the critical challenge of data scarcity in structural health monitoring (SHM) systems and introduces population-based SHM (PBSHM) as a solution. PBSHM allows for the transfer of information between similar structures to improve predictive models. The chapter discusses the risks associated with negative transfer in transfer learning and proposes a method to quantify the value of information transfer. By optimizing transfer strategies based on the expected value of information transfer, the chapter aims to minimize the risk of negative transfer, ensuring more accurate and reliable SHM systems. The methodology is demonstrated through a numerical case study, showcasing the potential of PBSHM in enhancing the safety and operational life of structures.AI Generated
This summary of the content was generated with the help of AI.
AbstractPopulation-based structural health monitoring (PBSHM) seeks to address some of the limitations associated with data scarcity that arise in traditional structural health monitoring (SHM). A tenet of the population-based approach to SHM is that information can be shared between sufficiently similar structures in order to improve predictive models. Transfer learning techniques, such as domain adaptation, have been shown to be a highly useful technology for sharing information between structures when developing statistical classifiers for PBSHM. Nonetheless, transfer learning techniques are not without their pitfalls. In some circumstances, for example, if the data distributions associated with the structures within a population are dissimilar, applying transfer learning methods can be detrimental to classification performance—this phenomenon is known as negative transfer. When considered in the context of operation and maintenance decision processes, negative transfer has significant implications. Deterioration in classification performance could translate to unnecessary inspections or repairs, and even critical maintenance interventions being missed entirely. Such changes in operation and maintenance strategy would result in additional costs being incurred and could undermine the integrity and safety of structures. Given the potentially severe consequences of negative transfer, it is prudent for engineers to ask the question “when, what, and how should one transfer between structures”.The current chapter aims to demonstrate a transfer-strategy decision process for a classification task for a population of simulated structures in the context of a representative SHM maintenance problem, supported by domain adaptation. The transfer decision framework is based on the concept of the expected value of information transfer. In order to compute the expected value of information transfer, predictions must be made regarding the classification (and decision performance) in the target domain following information transfer. In order to forecast the outcome of transfers, a probabilistic regression is used here to predict classification performance from a proxy for structural similarity based on the modal assurance criterion. -
Employing Guided Wave-Based Damage Localization Techniques for Additively Manufactured Plates with Different Infill Densities
Paweł H. Malinowski, Samir Mustapha, Mohammad Ali Fakih, Shishir Kumar SinghThe chapter delves into the application of guided wave-based damage localization techniques for additively manufactured plates with different infill densities. It begins by discussing the growing need for efficient nondestructive evaluation methods due to the increasing use of additively manufactured parts. The study focuses on poly(lactic acid) plates with varying infill densities, employing piezoelectric transducers and a scanning laser vibrometer to measure wave propagation. The authors introduce a data analysis approach that extracts specific signal parts to identify damage-reflected waves, which are then summed up to provide a final result. The results indicate that the infill density and excitation frequency significantly influence the localization algorithm's output. The study highlights the challenge of weak wave reflections in the investigated material, making localization difficult. Overall, this chapter provides valuable insights into the effectiveness of guided wave techniques for damage detection in additively manufactured materials.AI Generated
This summary of the content was generated with the help of AI.
AbstractThe additively manufactured (AM) parts have been introduced in many industrial fields, and their usage is growing. This increase of use forces the need for effective tools for damage evaluation. This study focuses on elastic guided waves. Their propagation was analyzed, and wave scattering was used for damage assessment. The samples under investigation are flat plates printed with poly(lactic acid) (PLA). A few plates were prepared with five infill densities. Solid printed parts use a lot of material, so the reduction of the infill density allows to save material required for manufacturing. On the other hand, the reduction of the infill density makes the structure similar to a sandwich composite, and such complex structure influences the guided wave propagation behavior. In this study, plates were prepared as healthy and with simulated damage. The simulated damage was a void introduced in the printing process. The guided waves in the plates were excited with surface-mounted piezoelectric transducers, while the sensing was realized with scanning laser Doppler vibrometer. Damage localization algorithms based on wave reflections were developed, and their performance was analyzed as a function of the infill density. The obtained results show a potential of guided waves-based techniques for the structural health monitoring of additively manufactured structures. -
Optimal Modeling of Deep Groove Ball Bearings for Application in Multibody Dynamics Simulations
Josef Koutsoupakis, Dimitrios GiagopoulosThe chapter explores the critical role of deep groove ball bearings in mechanical systems and the importance of real-time condition monitoring to maintain structural integrity. It delves into the challenges of traditional monitoring methods and introduces numerical simulations as a viable alternative for generating healthy and damaged state datasets. The authors develop a computationally efficient bearing model using an analytical approach and optimize it using AI methods to ensure high-fidelity data. This model is then used to simulate faults in bearings and train a deep learning classifier for damage detection. The chapter highlights the successful application of the model in accurately capturing the system's behavior and demonstrates its potential to improve numerical simulations and condition monitoring in various industries.AI Generated
This summary of the content was generated with the help of AI.
AbstractIn this work, optimal modeling of deep groove ball bearings is examined for application in multibody dynamics simulations. First, the equations for the normal contact force between the rolling elements and the two bearing races are established, including the effects of raceway surface roughness. Various contact force models with different hysteresis damping formulations are examined in order to select the best suited for the application. The bearing contact force model is then used in a multibody dynamics simulation of a bearing test-rig, aiming to estimate the model’s optimal parameters resulting in a good approximation of the system’s behavior and, finally, to a well-calibrated ready-to-use bearing model. The optimal model is then used to examine the system’s behavior in the presence of defects in the bearings, validating the robustness and performance of the optimized deep groove ball bearing model. The system’s response is examined by means of signal analysis as well as by using deep learning methods in order to characterize the health state of the system, thus proving the applicability of the present bearing modeling method for condition monitoring applications. -
Utilization of Bridge Acceleration Response for Indirect Strain Sensing
Soheila Sadeghi Eshkevari, Debarshi Sen, Soheil Sadeghi Eshkevari, Iman Dabbaghchian, Giulia Marasco, Shamim PakzadThe chapter delves into the utilization of bridge acceleration response for indirect strain sensing, addressing the challenges of traditional strain measurement methods. It introduces an innovative AI-driven framework that combines transformers and convolutional neural networks (CNNs) to estimate strain responses from bridge acceleration data. This approach is demonstrated through a case study on the Gene Hartzell Memorial Bridge in Easton, Pennsylvania, showcasing its effectiveness in both time history prediction and rainflow counting diagrams. The framework offers a cost-effective and robust solution for bridge fatigue assessment, enhancing the classic approaches in the field.AI Generated
This summary of the content was generated with the help of AI.
AbstractFatigue assessment in bridges is crucial for maintaining the American road infrastructure. Typically, this involves measuring the bridge’s strain response under daily traffic loading. However, deploying and maintaining strain sensors is costly and labor-intensive compared to acceleration sensors. To address this issue, we propose a neural network architecture that can perform indirect sensing by estimating strain from measured acceleration response. Our proposed network employs convolutional neural networks (CNN) and transformers to account for uncertainties and noise associated with field data, and effectively convert accelerations to strain by capturing both the pseudo-static and dynamic features of the strain response. To demonstrate the effectiveness of our framework, we use field data collected from the Gene Hartzell Memorial Bridge in Easton, Pennsylvania, USA, as a case study. With our novel approach, we can estimate strain with high accuracy from acceleration data and reconstruct rainflow cycle counting diagrams that can subsequently be used for bridge condition and life cycle assessment. -
Transfer Learning Across Heterogeneous Structures Through Adversarial Training
Mohammad Hesam Soleimani-Babakamali, Onur Avci, Serkan Kiranyaz, Ertugrul TacirogluThe chapter presents a groundbreaking approach to structural health monitoring (SHM) using transfer learning (TL) and adversarial training. It addresses the challenge of detecting structural damage across heterogeneous structures by employing full-spectrum FFT amplitudes as features. The proposed method involves training a deep neural network as a damage detector and using domain adaptation to transform target domain data into the source domain without the need for model retraining. This approach is tested on three benchmark datasets—Z24 bridge, Yellow Frame, and QUGS—demonstrating its effectiveness in distinguishing between undamaged and damaged cases. The chapter highlights the advantages of zero-shot learning and the potential for large-scale SHM applications, making it a valuable resource for professionals seeking innovative solutions in civil infrastructure monitoring.AI Generated
This summary of the content was generated with the help of AI.
AbstractTransfer learning (TL) methods have become increasingly crucial for the challenges in gathering accurately labeled data from various structures in structural health monitoring (SHM) tasks, such as structural damage detection (SDD). The structures must meet specific similitude criteria for the proposed TL technique’s effectiveness in current one-to-one domain approaches. To overcome this challenge, the authors have developed a novel TL method that utilizes raw vibrational features and raw-feature-to-raw-feature domain adaptation (DA) through spectral mapping. This approach offers a generalizable TL strategy that works across vastly different structures. The authors used generative adversarial network (GAN) architecture for the “learning,” as it can accommodate high-dimensional inputs in a zero-shot setting. The proposed TL approach was successfully evaluated over three structural health monitoring (SHM) benchmarks. Area under the curve (AUC) of the receiver operating characteristics (ROC) curve resulted in a threshold-bias-free estimation of SDD models retaining as much as 99% of the source model’s AUC through its application across different systems with diverse damage-representative data cases. -
Physics-Informed Machine Learning Part II: Applications in Structural Response Forecasting
Austin R. J. Downey, Eleonora Maria Tronci, Puja Chowdhury, Daniel CobleThe chapter delves into the innovative use of Physics-Informed Machine Learning (PIML) for forecasting structural responses, particularly in civil structures. It introduces the equation of motion as the core of structural modeling and explains how PIML integrates physical principles into machine learning algorithms to overcome the limitations of traditional approaches. The chapter presents a physics-constrained neural network method that enhances predictive accuracy by incorporating physical constraints into the training process. A numerical example of a cantilever beam model illustrates the significant improvement in structural response forecasting achieved by this method. The work concludes by emphasizing the transformative potential of PIML in structural modeling and monitoring, bridging the gap between theoretical physics and real-world applications.AI Generated
This summary of the content was generated with the help of AI.
AbstractPhysics-informed machine learning is a methodology that combines principles from physics with machine learning techniques to enhance the accuracy and interpretability of predictive models. By incorporating physical laws and constraints into the learning process, physics-informed machine learning enables more robust predictions and reduces the need for large amounts of training data. In part II of this two-part series, the authors present structural response forecasting using a physics-constrained methodology to solve the homogeneous second-order differential equations that constitute the equation of motion of a linear structural system. This forward problem is formulated to allow the incorporation of numerical methods into the training process while using segmented training to circumvent intrinsic stability limitations to the physics-informed machine learning problem. The ability of physics-informed machine learning to make generalizations for limited training data is discussed. -
Frequency-Based Damage Detection Using Drone-deployable Sensor Package with Edge Computing
Ryan Yount, Joud N. Satme, Austin R. J. DowneyThe chapter delves into the innovative application of drone-deployable sensor packages equipped with edge computing for structural health monitoring. It introduces an advanced UAV-deployable sensing node featuring electropermanent magnets and RF communication, designed for long-term deployment and real-time data analysis. The integration of an edge processor enables the computation of key features such as the fast Fourier transform (FFT) and power spectral density (PSD) directly on the sensor package, facilitating swift damage detection. The chapter highlights the power consumption profile of the sensor package, demonstrating the viability of edge processing for real-time structural analysis. Through practical tests and detailed analysis, the work underscores the potential of this technology for rapid assessment of infrastructure, particularly in challenging environments. The chapter concludes by discussing future improvements, including the use of smaller edge processors and networked sensor packages for comprehensive structural monitoring.AI Generated
This summary of the content was generated with the help of AI.
AbstractFor rapid infrastructure assessment following natural and man-made emergencies, the utilization of minimally invasive and cost-effective drone deployable sensor packages has gained significant attention. While compact sensors with wireless data transfer capabilities have demonstrated potential for monitoring structural dynamics of critical infrastructure, such systems typically require data to be processed off-device and often off-site. These additional steps hinder the rapid assessment aspect. A challenge arises when transmission is not feasible due to degraded communication links during natural or man-made emergencies. Moreover, off-site data processing may add unnecessary delays to actions that can be taken by emergency personnel following infrastructure damage. To maximize the usefulness of sensor packages for rapid infrastructure assessment, the integration of edge computing techniques into the sensors themselves to analyze data in real time presents a promising solution. The objective of this work is to demonstrate edge computing for frequency-based structural health monitoring techniques to showcase the effectiveness of on-device data processing for the rapid assessment of infrastructure. The proposed approach continuously computes the power spectral density of windowed vibration measurements taken from a structure of interest that has the potential to experience further damage, for example, the monitoring of a bridge immediately after a flooding event. This work presents contributions in terms of a methodology, focusing on the hardware implementation of edge computing algorithms. Additionally, a study of the performance and resource utilization of a windowed power spectral density processing algorithm on-device is provided. -
Understanding High-Frequency Modes in Electromechanical Impedance Measurement Using Noncontact Vibration Testing
Sourabh Sangle, William C. Rogers, Mohammad I. Albakri, Pablo A. TarazagaThis chapter delves into the understanding of high-frequency modes in electromechanical impedance (EMI) measurements using noncontact vibration testing. Traditional EMI measurements lack spatial information, making it challenging to understand the dynamic response of structures. The authors propose a method to correlate EMI measurements with the dynamic response of a structure, specifically focusing on an additively manufactured copper specimen. By comparing EMI measurements with mobility responses in the X-, Y-, and Z-directions, the chapter identifies modes that are sensitive to specific changes and quantifies the dominant directional dynamics. This approach offers a more informative analysis of structural health monitoring, paving the way for future work that includes extracting mode shapes and comparing damping and natural frequencies between EMI and dynamic responses.AI Generated
This summary of the content was generated with the help of AI.
AbstractElectromechanical impedance (EMI) measurements have been used for several decades in noninvasive health monitoring across various domains. Typical EMI measurements are recorded via a bonded piezoelectric transducer, at a high-frequency range, typically 30 kHz and above. Because EMI measurements are single input single output, the peaks in these measurements can be related to either mechanical, electrical, or coupled electromechanical modes, especially at higher frequencies. An attempt to move towards understanding these higher frequency modes is made in this study. To undertake this, noncontact vibration testing is carried out using a laser Doppler vibrometer (LDV). The specimen under investigation is bonded with a piezoelectric transducer with a prescribed voltage applied across the transducer. A comparison between the peaks in a recorded EMI measurement and the noncontact measurement is presented. A better understanding of these modes can be utilized for a better association between high-frequency measurement and physical attributes. Furthermore, tracking variations can lead to physical insights and mode transitions that more accurately inform material changes (damage). This work discusses the first step in this building association by exploring the relation between directional information from the noncontact measurement and coupled electromechanical information from the EMI measurement. This chapter also highlights some challenges and future work needed to make this technique robust and versatile for different specimens. -
On the Use of Symbolic Regression for Population-Based Modelling of Structures
G. Tsialiamanis, N. Dervilis, K. WordenThe chapter 'On the Use of Symbolic Regression for Population-Based Modelling of Structures' delves into the innovative use of symbolic regression for damage prognosis in structural health monitoring. It introduces the concept of population-based structural health monitoring, which aims to transfer knowledge from one structure to another, even when data is scarce. The application of symbolic regression is showcased through an experimental setup involving aluminium plates subjected to cyclic loading, where the algorithm successfully predicts damage evolution. The chapter concludes with promising results and outlines future directions to improve the method's accuracy and physical interpretability. This work is particularly notable for its practical approach to addressing real-world challenges in structural health monitoring, making it a must-read for professionals seeking cutting-edge solutions in the field.AI Generated
This summary of the content was generated with the help of AI.
AbstractModelling of structures is an important tool for decision-making regarding infrastructure. In the absence of sufficient knowledge of the physical phenomena that affect the structure, one can follow a data-driven approach to model its behaviour, relying exclusively on data acquired from it. However, a common problem of this approach is the scarcity of data or biased data. To deal with these two problems, approaches have been considered to transfer knowledge via machine learning models from one domain to another. The current work considers the case of population-based structural health monitoring (PBSHM) of structures. Such an approach is motivated by the common physics that dictates the behaviour of similar structures, which could offer an opportunity to exploit information from a population to create more robust and trustworthy models of data-poor structures of the same population. More specifically, the approach followed here is that of symbolic regression and the transfer is attempted between an extensively monitored structure and a data-poor structure for a regression application. The methodology is applied in a prognosis problem of crack growth in metal plates, and the results reveal the potential of symbolic regression to perform knowledge transfer. -
Markov Chain Monte Carlo on Matrix Manifolds for Probabilistic Model Order Reduction
Alessandra Vizzaccaro, Mikkel B. Lykkegaard, Tim DodwellThis chapter delves into the integration of Markov Chain Monte Carlo (MCMC) techniques with matrix manifolds for probabilistic model order reduction. The primary focus is on linear model order reduction (MOR), which aims to minimize the dimensionality of problems while preserving essential patterns. The methodology involves searching for optimal matrices on matrix manifolds, such as the Stiefel and Grassmann manifolds, to solve various applications including proper orthogonal decomposition, active subspaces, and matrix completion. By adapting MCMC algorithms like Metropolis-Hastings and Metropolis-adjusted Langevin algorithm, the work enables drawing samples from posterior distributions defined on matrix manifolds, effectively quantifying uncertainty. The proposed framework, implemented using the Manopt software, offers a robust approach to propagating uncertainty in reduced models, making it a valuable contribution to the field of probabilistic MOR.AI Generated
This summary of the content was generated with the help of AI.
AbstractThis work extends classical tools in linear model order reduction and more recent results in the field of optimization on Riemannian manifolds to the probabilistic case, within the Bayesian framework. We present a method to draw samples from a given target distribution defined on various matrix manifolds. The collected samples can be used to propagate uncertainty on the reduction matrix and other quantities of interest. -
Identification of Bird Species in Large Multi-channel Data Streams Using Distributed Acoustic Sensing
Andrew L. Jensen, William A. Redford, Nimran P. Shergill, Luke B. Beardslee, Carly M. DonahueThe chapter introduces Distributed Acoustic Sensing (DAS) as a promising technology for ecological monitoring, particularly for detecting bird calls. It discusses the advantages of DAS over traditional acoustic data acquisition methods, such as its ability to cover large areas with a single instrument. The chapter presents experiments conducted in different environments to test the effectiveness of DAS for bird call detection. It also explores various signal processing techniques, including short-term/long-term analysis and matched filtering, to optimize the detection and classification of bird calls. The chapter concludes by highlighting the potential of DAS for ecological monitoring but also acknowledges the challenges that need to be addressed for its widespread adoption.AI Generated
This summary of the content was generated with the help of AI.
AbstractThe health of an ecosystem can be challenging to monitor due to the complex nature of environmental systems. Fortunately, the health of a local ecosystem can be inferred by monitoring key species which are indicative of the overall health of the ecosystem. Microphones have emerged as a powerful tool for detecting bird calls of these key indicator species. However, using an array of microphones to monitor a large area requires a power source at each location in addition to sensor telemetry to retrieve the data. Distributed acoustic sensing (DAS) is a promising approach for large scale monitoring as a single hardware system is used to detect signals over large distances. We propose a novel application of DAS to detect avian species for ecological health monitoring. A single DAS interrogator unit and optical fiber can collect tens of kilometers of high frequency acoustic data with the added benefit that DAS does not suffer from time synchronization errors and remote power issues like traditional microphone arrays. This work investigates the performance of DAS when used to detect bird calls, with particular focus on the Great Horned Owl (GHO), an indicator species for prey vulnerability in an ecosystem. By quantifying the performance of several DAS configurations and bird call detection approaches, we demonstrate the potential of DAS for use in ecological health monitoring applications. -
A Machine Learning–Based Damage Estimation Model for Monitoring Reinforced Concrete Structures
Omair Inderyas, Sena Tayfur, Ninel Alver, F. Necati CatbasThe chapter discusses the development and evaluation of a machine learning-based damage estimation model for reinforced concrete structures using acoustic emission (AE) data. Acoustic emission is an effective technique for monitoring the progression of damage in structures, and various AE parameters such as amplitude, rise time, duration, and energy are crucial for determining the damage level. The study employs the K-nearest neighbors (KNN) algorithm, chosen for its speed and low computational cost, to classify damage types and degrees. The chapter presents three different damage classification models, each with varying numbers of damage classes, and evaluates their performance using metrics such as accuracy, precision, and recall. The results demonstrate that the two-class damage classification model outperforms the multiclass models, likely due to the curse of dimensionality. Additionally, the chapter highlights the importance of cumulative AE energy as the most significant feature for damage degree estimation. The study concludes that the KNN algorithm is a useful and effective tool for damage degree estimation in concrete structures, providing valuable insights for the civil engineering field.AI Generated
This summary of the content was generated with the help of AI.
AbstractAcoustic emission (AE) has gained much attention in recent years for its effectiveness in nondestructive testing and continuous monitoring of civil infrastructure. As a phenomenon, when structures are loaded, AE sensors mounted on the surface of the structure receive the waves released from the damage and convert them into electrical signals. By analyzing and evaluating these recorded signals, critical information such as type, location, time of origin, size, and orientation of the damage can be identified.Since the whole process, which includes a massive amount of dataset, is time-consuming and needs to be automated, this study aimed to develop a machine learning–based damage estimation model that would estimate fracture characteristics of concrete and would be a pioneer for the application of real-time and automated monitoring of structures. In this scope, failure behavior of concrete samples of various strengths and sizes under loading was monitored with AE using K-nearest neighbor (KNN) algorithm. A relationship was established between the load levels and damage status of the specimens with AE features. Afterward, the model was trained and tested, and fracture characteristic results estimated by the KNN models were evaluated to reveal its feasibility. -
Adaptive Radio Frequency Target Localization
Anthony A. Petrakian, Parker Segelhorst, Abigail Smith, Jeffery Dwayne Tippmann, Zigfried Hampel-AriasThe chapter delves into the critical tool of radio frequency (RF) localization, highlighting the limitations of traditional high-powered base stations and the need for more versatile, mobile solutions. It introduces a method using a single mobile sensing unit to localize RF signals through received signal strength (RSS) measurements, leveraging a particle filter and reinforcement learning. The study models the localization problem as a partially observable Markov decision process (POMDP), detailing the state, action, state transition function, observations, and rewards. The chapter also discusses the challenges of translating simulated models to real-world environments, addressing issues such as sensor variance and signal strength inaccuracies. Experiments are conducted to compare model performance and decision-making processes, demonstrating the potential of the proposed method in both restricted and optimal decision-making scenarios. The results show promising performance, particularly for models trained with moderately larger action distances. The chapter concludes by outlining future work directions, including the localization of dynamic targets, signal strength considerations, and the use of alternative technologies like drones for more efficient target localization.AI Generated
This summary of the content was generated with the help of AI.
AbstractMobile radio frequency (RF) target localization is a widely studied field with a wide variety of applications including health monitoring of electronic devices and search and rescue. This recently has been done using a greedy approach, where a sensor is moved closer to the target after each measurement. However, this does not allow for other constraints, such as maintaining a fixed distance from the target or optimization of energy consumption. Studies in the field have used techniques such as machine learning to localize static targets with received RF signals, and dynamic targets with received RF signals combined with line-of-sight observations. A recent study simulated localization of static and dynamic targets through RF signal characterization without the need for line-of-sight observation while also maintaining the previously mentioned constraints. This was done by modeling the problem as a Partially Observable Markov Decision Process (POMDP) and was solved through the use of particle filtering and reinforcement learning. The purpose of this work is to build upon this prior study by training a deep neural network in a simulated environment and applying inference in the real world. This led to various changes to the model that better matched real-world scenarios. By defining the metric of success as the distance between the actual location of the device and the estimated location of the device, it was shown that the model was able to accurately locate static devices within the measured standard deviation of the signal strength. Future work includes the use of autonomous units such as drones, as well as extending the capabilities of the model to localize real-world dynamic targets. -
Estimation of Acoustic Emission Arrival Time in Concrete Structures Using Convolutional Neural Network
Omair Inderyas, Ninel Alver, Aydin Kaya, Ulas BagciThe chapter delves into the application of a one-dimensional convolutional neural network (1D CNN) for accurately estimating the time of arrival (ToA) of acoustic emission signals in concrete structures. Traditional methods, such as short-term averaging (STA) and long-term averaging (LTA), often fall short in environments with low signal-to-noise ratios (SNR). The study compares the performance of 1D CNN models with different window sizes (300 and 1024 sample points) and varying numbers of convolutional layers. The results show that the 1D CNN model with two convolutional layers outperforms traditional methods, particularly in the window size of 300. This finding highlights the potential of 1D CNNs in enhancing the accuracy and reliability of damage localization in structural health monitoring applications.AI Generated
This summary of the content was generated with the help of AI.
AbstractAcoustic emission (AE) has recently gained a significant interest as a promising technique for monitoring damage progress in various structures including buildings, bridges, pipelines, and storage tanks. It relies on analyzing the acoustic activity, primarily associated with cracking phenomena, to assess structural integrity. However, one crucial parameter derived from AE signals is the time of arrival (ToA) of acoustic events, which is challenging to pick correctly. Accurate estimation of ToA is vital in localizing damage sources and enabling early detection of probable defects. Traditional approaches for ToA estimation often suffer from sensitivity to environmental and operational factors, such as imperfect coupling between AE transducers and the structures.To address this challenge, this study investigates the application of a one-dimensional convolutional neural network (1D CNN) for precise ToA estimation in AE signals. Experimental data acquired during compression tests on concrete specimens were utilized to train and test the model over windows of 300 and 1024 sample points in AE waveforms. By capturing a batch of representative acoustic features defined on a time basis and monitoring their evolution over time, the model was able to estimate ToA accurately in the window with 300 sample points. This proposed deep learning–based approach demonstrated promising potential for enhancing the accuracy and reliability of damage localization and early defect detection in various structural applications. -
Machine Learning–Based Method for Structural Damage Detection
Daniel Irawan, Evgeny V. Morozov, Murat TahtaliThe chapter delves into the application of machine learning, specifically convolutional neural networks (CNNs), for structural damage detection in thin-walled reinforced structures. It highlights the challenges posed by various types of damage in metallic and composite materials due to impacts and cyclic loading. The study presents a novel approach using a CNN ensemble network, trained with data from finite element method (FEM) simulations, to accurately locate and quantify damage in both isotropic and orthotropic plates. The method demonstrates superior performance in predicting damage parameters, showcasing significant improvements over single CNN models. The research also emphasizes the importance of ensemble networks in enhancing the accuracy of damage severity predictions, making it a valuable resource for professionals seeking advanced techniques in structural integrity assessment.AI Generated
This summary of the content was generated with the help of AI.
AbstractStructural damage assessment has become a necessity in the modern era. Researchers are trying to come up with better and faster ways to regulate structures’ health since traditional nondestructive testing (NDT) methods such as visual inspection take longer time to carry out and it is sensitive to the user’s skill in operating the apparatuses. Various methods incorporating reverse algorithms have been explored with the help of machine learning; however, we found out that the use of pretrained convolutional neural network (CNN) in detecting damage has not been explored widely. In this study, an ensemble network of CNNs is built based on GoogLeNet architecture to test its capability in detecting structural damages in plates. There are two cases of plate structure being tested for the model: isotropic plate (metallic structure) and orthotropic plate (composite structure). The damage induced in those plates is simulated with a reduction in mechanical properties, that is, elastic modulus in isotropic case and multidirectional elastic modulus in orthotropic case. The models try to pinpoint the location parameters of the damage in the plate and to quantify the severity of the damage itself by getting input variables from the modal properties of the plates. From the individual models, the information is then gathered using an ensemble network which is expected to improve the overall accuracy. The results from the final model show good correlation between predicted parameters and the actual case with promising results for further research. -
Correction to: Markov Chain Monte Carlo on Matrix Manifolds for Probabilistic Model Order Reduction
Alessandra Vizzaccaro, Mikkel B. Lykkegaard, Tim DodwellThe chapter offers a vital correction to the original publication on Markov Chain Monte Carlo on Matrix Manifolds for Probabilistic Model Order Reduction. The key update involves the accurate attribution of authorship to 'Mikkel B. Lykkegaard'. This correction is essential for maintaining the integrity of academic records and ensuring that credit is appropriately assigned. The chapter is part of the broader 'Data Science in Engineering Vol. 10' collection, emphasizing the importance of precise and up-to-date information in the field. By addressing this correction, the chapter highlights the significance of accurate author attribution in scholarly works, contributing to the overall reliability of scientific literature.AI Generated
This summary of the content was generated with the help of AI.
- Title
- Data Science in Engineering Vol. 10
- Editors
-
Thomas Matarazzo
François Hemez
Eleonora Maria Tronci
Austin Downey
- Copyright Year
- 2025
- Publisher
- Springer Nature Switzerland
- Electronic ISBN
- 978-3-031-68142-4
- Print ISBN
- 978-3-031-68141-7
- DOI
- https://doi.org/10.1007/978-3-031-68142-4
PDF files of this book don't fully comply with PDF/UA standards, but do feature limited screen reader support, described non-text content (images, graphs), bookmarks for easy navigation and searchable, selectable text. Users of assistive technologies may experience difficulty navigating or interpreting content in this document. We recognize the importance of accessibility, and we welcome queries about accessibility for any of our products. If you have a question or an access need, please get in touch with us at accessibilitysupport@springernature.com
