Health Monitoring Approach Research Articles

Acoustic emission entropy: An innovative approach for structural health monitoring of fracture‐critical metallic components subjected to fatigue loading

AbstractThe paper presents an innovative approach for structural health monitoring of metallic components under fatigue crack phenomena. The methodology is based on the evaluation of the information entropy of the acoustic emission (AE) data. AE testing of fatigue crack growth (FCG) is performed on metallic components is performed within an extremely noisy testing environment. Basic AE data analysis is demonstrated to be inefficient with regard to the specific testing conditions. AE entropy is proven to be a reliable damage‐sensitive feature for real‐time assessment despite both significant noise disturbance and complexity/randomness of the acoustic phenomena. This was also confirmed for (time‐)discontinuous monitoring processes over random‐based data detections. An innovative monitoring protocol is finally developed according to the experimental evidence also considering the recommendations of the current monitoring. The protocol is found to be promising for structural health monitoring of metallic fracture‐critical components of structures under fatigue.

Data quality indicators for vibration-based damage detection and localization

The effective application of vibration-based damage detection (VBDD) methods as a structural health monitoring approach depends largely on the accurate measurement of modal properties, particularly the mode shapes. However, the modal properties of bridges and other civil engineering structures are commonly measured using an output-only approach and ambient excitation sources, which can lead to considerable variability in the measurements and therefore a lack of confidence in the reliability of the VBDD results. This paper proposes two data quality indicators that can be used to assess the quality of a calculated VBDD parameter in terms of its potential to confidently identify and characterize damage, based on the consistency of the parameter when obtained from different sets of vibration test data. The performance of the indicators is demonstrated by application to a simple-span slab-on-girder bridge deck and comparison to the probabilities of successfully locating nine damage cases, as calculated using transient dynamic finite element analyses under simulated randomly varying loads. The data quality indicators were found to correlate well with the probability of successful localization, and were able to reflect the varying probabilities associated with several factors, including the number of repeated random trials used to estimate the mode shapes, the distance from the damage to the nearest sensor, the proximity of the damage to simple supports, and the severity of the damage. The proposed parameters were therefore shown to be capable of determining whether the available data are of sufficient quality to confidently apply VBDD methods.

A Feasibility Study on Monitoring Pile‐Soil Bonding Condition Using Piezoceramic Transducer and Horizontal Impact

The in situ evaluation of pile‐soil bonding condition plays an important role for pile safety assessment in its life cycle. However, so far, there is still no fully mature tool to analyze such couplings, since the pile‐soil coupling exhibits complex and time‐varying relationships. This paper innovatively proposes a health monitoring approach to evaluate the bonding status of the soil and pile contact area. An impact method based on a piezoelectric ceramic sensor is proposed to monitor the bond of pile and soil. A horizontal impact was introduced near the top of the pile, and the induced stress waves were detected by the piezoceramic smart aggregate (SA) sensor embedded in the pile. Different crack damage sizes were made between the soil and the pile to investigate the change of the bonding. An energy index was developed to quantitatively evaluate the quality of the bonding as a pile‐soil bonding index. The proposed approach inspired a potential way to directly judge if there is crack damage between the pile and soil and to evaluate pile safety.

FedHome: Cloud-Edge Based Personalized Federated Learning for In-Home Health Monitoring

In-home health monitoring has attracted great attention for the ageing population worldwide. With the abundant user health data accessed by Internet of Things (IoT) devices and recent development in machine learning, smart healthcare has seen many successful stories. However, existing approaches for in-home health monitoring do not pay sufficient attention to user data privacy and thus are far from being ready for large-scale practical deployment. In this paper, we propose FedHome, a novel cloud-edge based federated learning framework for in-home health monitoring, which learns a shared global model in the cloud from multiple homes at the network edges and achieves data privacy protection by keeping user data locally. To cope with the imbalanced and non-IID distribution inherent in user’s monitoring data, we design a generative convolutional autoencoder (GCAE), which aims to achieve accurate and personalized health monitoring by refining the model with a generated class-balanced dataset from user’s personal data. Besides, GCAE is lightweight to transfer between the cloud and edges, which is useful to reduce the communication cost of federated learning in FedHome. Extensive experiments based on realistic human activity recognition data traces corroborate that FedHome significantly outperforms existing widely-adopted methods.

Prognostics and Health Monitoring Methodologies and Approaches: A Review

Prognostics is a term that engineering borrowed from medicine to refer to the discipline concerned with the Remaining Useful Life (RUL) of an engineering device. This paper surveys the RUL prediction techniques and classifies them into four categories of model-based techniques,knowledge-based techniques, experience-based techniques, and data-driven techniques. A comparative review is given for the main features, prominent advantages, potential shortcomings and main subcategories for each of these categories. The survey is supported by an extensive listfor up-to-date references.

Pulsed terahertz spectroscopy combined with hybrid machine learning approaches for structural health monitoring of multilayer thermal barrier coatings.

Structural health monitoring of multilayer thermal barrier coatings (TBCs) is very vital to ensure the structural integrity and service performance of the hot-section components of the aero-engine. In this paper, we theoretically and numerically demonstrated that the terahertz time domain spectrum and the terahertz reflectance spectrum could be adopted to estimate the structure parameters, based on the finite difference time domain (FDTD) algorithm, 64 samples which were imported with three kinds of 64 sets structure parameters had been calculated to obtain the time domain and terahertz reflectance signals. To mimic the actual test signals, the original FDTD simulation signals were processed by adding the Gaussian white noise and wavelet noise reduction. To reduce the data dimension and improve the calculation efficiency during modeling, the principal component analysis (PCA) algorithm was adopted to reduce the dimensions of time-domain data and reflectance data. Finally, these data after multiple signal processing and PCA feature extraction were used to train the extreme learning machine (ELM), combining the genetic algorithm (GA) could optimize the PCA-ELM model and further improve the prediction performance of the hybrid model. Our proposed novel and efficient terahertz nondestructive technology combined with the hybrid machine learning approaches provides great potential applications on the multilayer TBCs structural integrity evaluation.

A methodological approach for structural health monitoring of mass-timber buildings under construction

Structural health monitoring (SHM) is a method used to evaluate the performance of new structural systems and critical infrastructure. With mass-timber building construction on the rise, SHM programs have emerged to document hygrothermal, static, and dynamic behavior of these structures. To most efficiently document behavior and provide recommendations to industry, it is key that the research community work collaboratively to create consistent data by using standardized approaches. This paper presents a methodological approach for monitoring mass-timber buildings during construction to address this need. The approach was validated over ten months with a mass-timber building under construction at Oregon State University.

A Fuzzy Krill Herd Approach for Structural Health Monitoring of Bridges using Operational Modal Analysis

Monitoring bridges is an important issue in the structural safety assessment. As a consequence of inaccessible sections of these structures, the utilization of nondestructive damage identification techniques seems to be vital for maintaining safety and integrity. In this study, a new hybrid Fuzzy Krill Herd approach is performed based on online responses to assess the global structural integrity of an in-service bridge. While the algorithms have been studied previously in the literature, their efficiency in structural health monitoring of a real bridge structure has yet to be investigated. The methodology is applied to two types of steel girders and concrete bridges in this manuscript. The finite element (FE) modeling is considered for further numerical investigation of dynamic characteristics and structural behavior of the bridges. To evaluate the efficiency of the proposed technique, a three-dimensional FE model and a developed simple two-dimensional girders models are simulated. The results indicate the capability of the fuzzy logic approach in obtaining accurate information in the presence of noisy input data or the data with missing values. Based on the results, increasing the number of the measurement modes and using the torsional modes lead to an accurate damage diagnosis process even in symmetric structures.

Quantification of Multi-Use Trail Effects Using a Rangeland Health Monitoring Approach and Google Earth

Creation and use of multi-use trails are increasing throughout grasslands of North America. While the direct and indirect ecological impacts of multi-use trails are generally understood, their specific impacts on adjacent grassland conservation require further assessment. Traditional scientific methods of quantifying trail impacts are often prohibitively costly in terms of required time, expertise, and equipment. Here, we evaluate the utility of a rapid assessment methodology—combining rangeland health protocols for grasslands with publicly available Google Earth mapping technologies—for capturing trail impacts as a function of distance from trail in a multi-use natural area in southwestern Alberta, Canada. Our methodology successfully detected a positive relationship between rangeland health scores and increasing distance from trail, indicating its viability as a rapid assessment tool. Second, this methodology was sensitive enough to allow the development of a more generalized statistical model demonstrating that rangeland health was best explained by a combination of slope, aspect, plant community type, and distance from trail. Combined, we suggest the limited costs of this method, combined with its ability to detect indirect impacts of trails on the health of adjacent grasslands, indicate this tool has potential utility for land managers where resources are limited. More specifically, we suggest this grassland health protocol can be highly effective as a first “rapid assessment,” prior to investing in more traditional ecological methodologies.Jessica Grenke, James F. Cahill, and Edward W. Bork. 2018. “Quantification of Multi-Use Trail Effects Using a Rangeland Health Monitoring Approach and Google Earth.” Natural Areas Journal 38(5):370-379.

21.K. Workshop: Closing the data gap on child health – approaches, characteristics and methodological challenges

Abstract As stated in the 2010 and 2020 Marmot review, inequalities, especially during childhood, have lifelong impacts. Social hardships during childhood reduce the chances for children to develop and flourish and show to have a negative impact on livelong health outcomes. To monitor progress in reducing inequalities in child health and to inform policy makers and other stakeholders to develop healthy public policies, data on social determinants, health outcomes, health behaviour, the conditions in which children grow up, and the utilisation of health care are needed. But these data have been scarce for routine population health monitoring. Routine data sources like mortality statistics or hospital discharge data have shown to be not very well suited to monitor child health since they are only outcome focussed and cover only extreme events. In this workshop we aim to showcase different national approaches that aim to overcome this data gap. We want to discuss their set-up including the strength and weaknesses in terms of organisation, management, data quality, information richness, and dissemination and we want to learn about the experiences workshop participants made in their countries. First, we will learn about two approaches that combine questionnaire and examination elements. England has long-standing experiences in monitoring child health and includes children in their general population-based health examination survey, so that children's health can be seen in the context of other household members, in particular parent's socio-economic status, health status, and health-related behaviours. The German approach focusses specifically on child health and includes a module for longitudinal monitoring. Second, we will learn about experiences from Finland who make use of routinely collected data from health check-ups performed at child health clinics and school health care. While the first two approaches use representative population samples, the Finish approach make use of a whole population sample including those who utilise public health services. The third part of this workshop will focus on a specific methodological discussion regarding the identification of children with chronic diseases and the comparison of several measurement approaches. We will devote twenty minutes for the presentation of the three different data collection approaches in order to gain a good overview on their set-up and their characteristics and 15 minutes for the methodological presentation regarding the measurement of chronic diseases in children. 15 minutes are reserved for further discussion with the workshop participants and the exchange of experiences in child health monitoring worldwide. Key messages Register- and survey-based approaches for child health monitoring exist that provide data for routine population health monitoring and inform policy-making. Methodological issues, e.g. in terms of standards or measurement approaches exist and need further research and discussion.

A Digital Twin approach based on nonparametric Bayesian network for complex system health monitoring

This paper proposes a Digital Twin approach for health monitoring. In this approach, a Digital Twin model based on nonparametric Bayesian network is constructed to denote the dynamic degradation process of health state and the propagation of epistemic uncertainty. Then, a real-time model updating strategy based on improved Gaussian particle filter (GPF) and Dirichlet process mixture model (DPMM) is presented to enhance the model adaptability. On one hand, for those parameters in the nonparametric Bayesian network with prior models, the improved GPF is used to update them in real time. On the other hand, for parameters lacking a prior model, DPMM is proposed to learn hidden variables, which adaptively update the model structure and greatly reduce uncertainty. Experiments on the electro-optical system are conducted to validate the feasibility of the Digital Twin approach and verify the effectiveness of the nonparametric Bayesian network. The results of comparative experiments prove that the Digital Twin approach based on nonparametric Bayesian Network has a good model self-learning ability, which improves the accuracy of health monitoring.

A Digital Twin-Based Approach for the Fault Diagnosis and Health Monitoring of a Complex Satellite System

The ever-increasing functional density and complexity of the satellite systems, the harsh space flight environment, as well as the cost reduction measures that require less operator involvement are increasingly driving the need to develop new approaches for fault diagnosis and health monitoring (FD-HM). The data-driven FD-HM approaches use signal processing or data mining to obtain implicit information for the operating state of the system, which is good at monitoring systems extensively and shallowly and is expected to reduce the workload of the operators. However, these approaches for the FD-HM of the satellite system are driven primarily by the historical data and some static physical data, with little consideration for the simulation data, real-time data, and data fusion between the two, so it is not fully competent for the real-time monitoring and maintenance of the satellite in orbit. To ensure the reliable operation of the complex satellite systems, this paper presents a new physical–virtual convergence approach, digital twin, for FD-HM. Moreover, we present an FD-HM application of the satellite power system to demonstrate the effectiveness of the proposed approach.

Nutzung der Haptoglobinkonzentration im Blutserum als Indikator im Tiergesundheitsmonitoring bei Milchkühen

To investigate the association between haptoglobin concentration in the blood and the occurrence of ketosis, selected clinical parameters as well as lameness in dairy cows. The data was collected on 39 dairy farms in Bavaria over a period of 8 months. In 712 Simmental and Brown Swiss cows, clinical examinations as well as milk and blood samplings were performed between 10 and 30 days after calving. In these blood samples, the concentrations of non-esterified fatty acids (NEFA), β-hydroxybutyric acid (BHBA) and haptoglobin (Hp) were determined. Analysis of the milk included milk constituents (fat, protein, urea, lactose and acetone), BHBA, NEFA and the somatic cell count (SCC). Significant correlations were found between increased Hp-concentration on the one hand and increased NEFA levels in blood and milk (p < 0.001), increased somatic cell count (p < 0.001), lameness (p < 0.001), as well as reduced lactose content (p < 0.001) and protein content in the milk (p = 0.001) on the other hand. Animals sampled during the warmer summer months showed significantly higher serum Hp-concentrations (p < 0,001). Heifers exhibited significantly higher Hp-values than multiparous individuals (p < 0.001). By dividing the examined cows into 4 clusters, a Hp-threshold value could be determined at 0.18 mg/ml. Combined with a SCC threshold of 40 500 cells/ml milk, the majority of animals with subclinical and clinical abnormalities could be identified. Measurement of the Hp-concentration in blood is a pertinent approach in animal health monitoring during the postpartum period. In combination with evaluations of milk amount and contents, deviations from the physiological status may be recognized and affected individuals treated early on. Haptoglobin may be used to assess the health status of the individual animal as well as an indicator of herd health in the context of animal health monitoring.

A conceptual study of utilizing compressive-sensing-based fan noise mode detection for aeroengine prognostic and health management.

In this paper, a concept about the aeroengine fan health monitoring approach is proposed based on the compressive-sensing-based acoustic mode detection method. Utilizing only a few acoustic sensors, possible accidents inducing the change of the fan noise mode spectrum can be inferred. To enable such a concept, the array design strategy and optimization method are first studied by maximizing the incoherence of the so-called sensing matrix. The performance of the designed array is examined in both simulation and experimental studies. Then, the idea of fan noise monitoring is conceptually demonstrated in wind tunnel tests by taking into account possible accidental scenarios with foreign body intrusions. The simulation and experimental results suggest that under such circumstances remarkable changes appear in the azimuthal mode spectrum from fan noise. Finally, it is demonstrated that the fan noise variation can be successfully detected by the compressive sensing method with just six sensors. In this way, the foreign body intrusion can be further diagnosed through the combination of compressive sensing and mode detection. Overall, the results confirm the potential capability of the proposed concept for future aeroengine health monitoring applications.

Benchmarking Deep Neural Network Architectures for Machining Tool Anomaly Detection

Abstract With the democratization of cyber-physical systems, edge computing, and large-scale data infrastructure, the volume of operational data available is continuously increasing. One of the significant challenges in current industrial research is defining a robust and scalable approach for machine health monitoring and anomaly detection. The methods that exist for such purposes rely extensively on feature engineering and are strongly dependent on the expertise of the operator, hence limiting their generalization. Deep learning techniques, on the other hand, are known to automate feature engineering and allow complex abstractions to be learned, making them particularly suitable for machine health monitoring. This paper presents a benchmarking of deep neural network architectures for the identification of machining tool anomalies on a lathe machine. The features are generated using indirect metrics such as sensor data and process variables from the machine controller, but without direct metrics like surface roughness or finished part quality. The ability of different architectures to identify incipient anomalies in tool quality is compared. A detailed treatment of various subsets of features is provided along with their relative importance to identify the minimum required parameters for accurately identifying the tool anomaly for each architecture considered. Finally, a recommendation is provided based on the results obtained on the type of architecture that is appropriate for the identification of machining tool anomalies. It is expected that the embeddings of the training data learned by the chosen network can be used for other learning tasks, such as transfer learning to another machine or anomaly type. The methodology described in this paper lends itself well to continuous monitoring through the use of scalable robust models and appropriate units of analysis for each model.

Health monitoring of bolted joints using modal-based vibrothermography

This article presents a novel modal-based vibrothermographic approach for health monitoring of loosening bolted joints in coupled structures. In this article, the theoretical background supporting this proposed approach is firstly presented. Through finite element analyses on a simple bolted structure with varying joint conditions achieved by adjustment of bolt loads, the relationship between the bolt load and the temperature increase in the vibrating bolted joint during vibrothermographic tests was revealed. Experimental vibrothermographic tests on a more complex structure were completed to verify the observations from the finite element analyses while demonstrating the viability of the vibrothermographic approach in a laboratory environment. It has been shown that this vibrothermographic approach was able to determine the stage of a bolted joint in its progression of failure by tracing the changes in the temperature increase in relevant regions during vibrothermographic tests. Moreover, additional tests have been performed to illustrate that this approach was effective even by using only the residual responses of the structure’s vibration that were away from the resonances, which indicates it is more applicable to structures with higher damping as such structures have stronger residual responses during vibration that can be utilized. In the concluding observations of the article, the procedure for practical application of this approach is summarized, and its potential for further development is discussed.

Vibration Characterization of the Human Knee Joint in Audible Frequencies.

Injuries and disorders affecting the knee joint are very common in athletes and older individuals. Passive and active vibration methods, such as acoustic emissions and modal analysis, are extensively used in both industry and the medical field to diagnose structural faults and disorders. To maximize the diagnostic potential of such vibration methods for knee injuries and disorders, a better understanding of the vibroacoustic characteristics of the knee must be developed. In this study, the linearity and vibration transmissibility of the human knee were investigated based on measurements collected on healthy subjects. Different subjects exhibit a substantially different transmissibility behavior due to variances in subject-specific knee structures. Moreover, the vibration behaviors of various subjects’ knees at different leg positions were compared. Variation in sagittal-plane knee angle alters the transmissibility of the joint, while the overall shape of the transmissibility diagrams remains similar. The results demonstrate that an adjusted stimulation signal at frequencies higher than 3 kHz has the potential to be employed in diagnostic applications that are related to knee joint health. This work can pave the way for future studies aimed at employing acoustic emission and modal analysis approaches for knee health monitoring outside of clinical settings, such as for field-deployable diagnostics.

Sleep Monitoring With Hidden Markov Model for Physical Conditions Tracking

Sleep monitoring is an important health monitoring approach. Generally, one's physical conditions fall into several qualities such as normal, weak, etc. And the clues of physical conditions are buried in the mass monitoring data. In this paper, a physical condition tracking method is proposed based on sleep monitoring with Hidden Markov Model (HMM). Firstly, physical features are acquired during sleep over a fixed time period using a micro movement sensitive mattress sleep monitoring system (MSMSMS). A feature matrix is created and used to derive an HMM that represents the condition of the subject. The connection between the variation of physical conditions and the Markov process is established. Bayesian information criterion is utilized to determine the number of hidden states in an HMM. Once a reference HMM is established, long-term physical conditions variation tracking can be realized by evaluating the fitnesses between upcoming monitoring data and the reference HMM. The proposed method is verified on a sleep monitoring experiment using the MSMSMS.

Impedance-based structural health monitoring applied to steel fiber-reinforced concrete structures

The goal of the present contribution is to test the effectiveness of an impedance-based structural health monitoring (ISHM) methodology to detect damage in steel fiber-reinforced concrete structures. For this aim, a piezoceramic sensor coated with epoxy is used. A prismatic concrete specimen was evaluated for the case in which the damage condition was obtained from compression testing. The impedance signatures were measured both before and after the compression testing for the specimens placed inside an environmental chamber. An optimization procedure was applied to compensate for the temperature effects that appear on the impedance signatures. For illustration purposes, the coated piezoelectric transducer was initially bonded to the surface of an aluminum beam to detect damage induced by a loosening bolt. An impedance analyzer was used to measure impedance signatures in the frequency range of 40–70 kHz for the procedure with the aluminum beam and 50–80 kHz for the prismatic steel fiber concrete specimen. An in-house portable impedance meter (SySHM impedance meter) was used to measure the impedance signatures. The results encourage the use of the ISHM approach for health monitoring of steel fiber-reinforced concrete structures.

Acousto-ultrasonics-based health monitoring for nano-engineered composites using a dispersive graphene-networked sensing system

Sensing is a fundamental yet crucial part of a functional structural health monitoring system. Substantial research has been invested in developing new sensing techniques to enhance sensing efficiency and accuracy. Practical applications of structural health monitoring approaches to real engineering structures require strict criteria for the sensing system (e.g. weight, position, intrusion and endurance), which challenge existing sensing techniques. The boom in nanotechnology has offered promising solutions for the development of new sensing approaches. However, a bottleneck still exists when considering the density of sensors and surface-mounted modality of installation. In this study, graphene nanoparticles are dispersed into a glass fibre/epoxy composite to form a dispersive network sensing system. The piezoresistivity of the graphene-formed network changes locally as a result of the change of inter-nanoparticle distances which triggers the ‘tunnelling effect’ and drives the sensor to respond to propagating elastic waves. Due to the dense graphene network formed within the composite, only a small area is required, functioning as a single sensing element to capture ultrasonic waves. To validate such capability, passive acoustic emission tests and active guided ultrasonic wave tests are performed individually. The graphene-networked sensing system can precisely capture wave signals which contain effective features to identify impact spot or damage location. Integrating passive graphene-formed network and active lead zirconate titanate wafers can form a dense network, capable of fulfilling general structural health monitoring tasks.