arXiv subjects

Philipp Wittenberg

Publications and source records attributed to Philipp Wittenberg.

14 recordsLinked to original sources

arXiv2026

Removal of Multivariate Environmental Influences in Structural Health Monitoring through Conditional Covariances and Supervised Learning

In structural health monitoring (SHM) systems, data is collected from a multitude of sensors measuring, for example, vibration or strain in the structure, along with additional features that capture environmental or operational information. It is well known that changes in the measured sensor outputs do not necessarily originate from structural damage but are often induced by environmental changes. One popular approach to account for these effects is regressing the system outputs on the confounding factors, also known as "response surface modeling". Afterward, the predicted values are subtracted from the observed ones to obtain corrected data with the environmental effects (supposedly) removed. However, the evaluation of real-world SHM data shows that environmental conditions may affect not only the expected output values but also higher-order statistical moments, particularly the variances of and the covariances and correlations between the output quantities, such as eigenfrequencies of different modes or strain sensors at different locations. By construction, the (supervised) machine learning techniques commonly used for response surface modeling cannot account for those higher-order effects. To address these issues, we present and discuss several approaches for identifying and quantifying multivariate confounding effects on output covariances and correlations: a nonparametric, kernel-based estimator, a random forest, a semiparametric additive model, and a deep learning approach. Furthermore, we show how the resulting conditional covariance matrices can be used in an SHM pipeline. We compare the competing methods on both artificial data and real-world load test data from the Vahrendorfer Stadtweg bridge in Hamburg, Germany, as well as eigenfrequency data from the railway bridge KW51 near Leuven, Belgium.

stat.ME↗

arXiv2026

Benchmarking Sensor-Fault Robustness in Forecasting

Cyber-physical system (CPS) forecasting models depend on sensor streams with noisy, biased, missing, or temporally misaligned readings, yet standard forecasting evaluation often selects models by nominal error without showing whether they remain robust under such faults. We introduce SensorFault-Bench, a shared CPS-grounded sensor-fault stress-test protocol for evaluating forecasting architectures and robustness-improvement methods, and an operational taxonomy organizing the method comparison. Across four real-world datasets and eight scored scenarios governed by a standardized severity model, it reports worst-scenario degradation, clean mean squared error (MSE), and worst-scenario fault-time MSE, separating relative robustness from absolute error. A disjoint fault-transfer split lets explicit fault-training methods train on adjacent fault families while evaluation uses separate benchmark scenarios. Empirically, forecasting architectures favored by clean MSE can degrade sharply under faults, and clean-MSE rankings can disagree with worst-scenario fault-time error rankings. Chronos-2, the evaluated zero-shot foundation-model representative, matches or trails the last-value naive forecaster in clean MSE on the two single-target datasets and has the largest worst-scenario degradation on ETTh1 and Traffic, where all channels are forecast targets. For the evaluated robustness-improvement method set, paired deltas show selective degradation reductions: projected gradient descent adversarial training and randomized training lead where value faults dominate observed degradation, while fault augmentation leads where availability faults dominate. SensorFault-Bench provides open-source code, documented data access, and reproduction and extension guides, so new datasets, architectures, and robustness-improvement methods can be evaluated under the same CPS sensor-fault robustness protocol.

cs.LG↗

arXiv2026

Feature Reconstruction and Monitoring of Load Test Data under Varying Environmental Conditions

System outputs in Structural Health Monitoring (SHM), such as sensor measurements or extracted features like eigenfrequencies, are influenced not only by (potential) damage but also by environmental and operational variables (EOV). Identifying these factors and removing their effects from the data is essential before proceeding with further analysis. Most existing methods for this task focus on the expected values of system outputs, e.g., using different types of response surface modeling. However, it has been shown that confounding variables can also affect the (co-)variance of and between system outputs. This is particularly important because the covariance matrix is an essential building block in many damage detection methods in SHM. Beyond standard response surface modeling, a nonparametric kernel approach can be used to estimate a conditional covariance matrix that can change depending on the identified confounding factor. This improves our understanding of how, e.g., temperature affects the system outputs. In this work, we present a new confounder-adjusted version of feature reconstruction. It uses the conditional covariance matrix as the basis for (conditional) principal component analysis. The resulting (conditional) principal component scores are then used to reconstruct system outputs with the confounding influences removed. In particular, the new approach eliminates the confounders effect on both the mean and the covariance. As will be shown on load test data from the Vahrendorfer Stadtweg bridge in Hamburg, Germany, the reconstructed features can then be employed for monitoring, e.g., using an appropriate control chart, resulting in fewer false alarms and a higher probability of detecting damage.

stat.AP↗

arXiv2026

Higher-Order Multivariate Environmental Influences in Structural Health Monitoring

System outputs such as eigenfrequencies or strain data, often used in structural health monitoring (SHM), not only react to damage but also depend on environmental conditions. When trying to correct for these confounding effects, it is often (at least implicitly) assumed that only the expected, i.e., mean, output values are affected by environmental conditions. However, the evaluation of real-world SHM data indicates that environmental conditions may influence not only the mean output but also higher-order statistical moments, particularly the variances of and the covariances and correlations between the output quantities, such as eigenfrequencies of different modes or strain sensors at different locations. To address these issues, we discuss two approaches for identifying and quantifying multivariate confounding effects on output covariances and correlations: a random forest and a nonparametric, kernel-based approach. We compare the two competing methods on both artificial and real-world SHM data, finding that the kernel-based approach achieves higher accuracy, but the random forest produces estimates that are more robust and sometimes easier to interpret.

stat.AP↗

arXiv2026

Covariate-Dependent Functional Principal Component Analysis for SHM

In Structural Health Monitoring (SHM), sensor measurements and derived features such as eigenfrequencies often exhibit systematic daily patterns and can therefore be naturally represented as functional data. Furthermore, these patterns are typically influenced by environmental factors, particularly temperature, which can substantially affect the observed system response. While most existing methods for removing environmental effects assume that confounding influences affect only the mean response, it has been shown that environmental and operational factors may also alter the covariance structure of the residual process. To address this limitation in a functional data monitoring framework, we incorporate so-called covariate-dependent functional principal component analysis (CD-FPCA), which allows eigenfunctions and eigenvalues of the residual process to vary smoothly with covariates such as temperature. The proposed methodology is illustrated using an extended version of the KW51 railway bridge eigenfrequency dataset. This case study suggests that accounting for covariate effects beyond the functional mean can improve the robustness of the monitoring procedure, in particular by reducing environmentally induced (false) alarms under challenging low-temperature conditions.

stat.ME↗

arXiv2025

Confidence Intervals for Conditional Covariances of Natural Frequencies

In structural health monitoring (SHM), sensor measurements are collected, and damage-sensitive features such as natural frequencies are extracted for damage detection. However, these features depend not only on damage but are also influenced by various confounding factors, including environmental conditions and operational parameters. These factors must be identified, and their effects must be removed before further analysis. However, it has been shown that confounding variables may influence the mean and the covariance of the extracted features. This is particularly significant since the covariance is an essential building block in many damage detection tools. To account for the complex relationships resulting from the confounding factors, a nonparametric kernel approach can be used to estimate a conditional covariance matrix. By doing so, the covariance matrix is allowed to change depending on the identified confounding factor, thus providing a clearer understanding of how, for example, temperature influences the extracted features. This paper presents two bootstrap-based methods for obtaining confidence intervals for the conditional covariances, providing a way to quantify the uncertainty associated with the conditional covariance estimator. A proof-of-concept Monte Carlo study compares the two bootstrap versions proposed and evaluates their effectiveness. Finally, the methods are applied to the natural frequency data of the KW51 railway bridge near Leuven, Belgium. This real-world application highlights the practical implications of the findings. It underscores the importance of accurately accounting for confounding factors to generate more reliable diagnostic values with fewer false alarms.

stat.AP↗

arXiv2025

Multivariate Long-term Profile Monitoring with Application to the KW51 Railway Bridge

Structural Health Monitoring (SHM) plays a pivotal role in modern civil engineering, providing critical insights into the health and integrity of infrastructure systems. This work presents a novel multivariate long-term profile monitoring approach to eliminate fluctuations in the measured response quantities, e.g., caused by environmental influences or measurement error. Our methodology addresses critical challenges in SHM and combines supervised methods with unsupervised, principal component analysis-based approaches in a single overarching framework, offering both flexibility and robustness in handling real-world large and/or sparse sensor data streams. We propose a function-on-function regression framework, which leverages functional data analysis for multivariate sensor data and integrates nonlinear modeling techniques, mitigating covariate-induced variations that can obscure structural changes.

stat.AP↗

arXiv2024

Data Set of Load Tests and Structural Health Monitoring of a concrete Box Girder bridge

Static and dynamic load tests were conducted on an anonymized in-service prestressed concrete box girder bridge constructed in 1972 and designed for Bridge Class~30 according to DIN~1072. The bridge is instrumented as part of the DTEC-SHM research initiative with a permanent structural health monitoring system comprising 128 sensors, including accelerometers, strain sensors, inclinometers, displacement transducers, embedded temperature sensors, and weather stations. The test program includes long-term static loading with stepwise added masses of 680 kg, 1420 kg, and 2160 kg, short-term static truck parking positions, and repeated dynamic truck crossings under loaded and unloaded vehicle configurations. The short-term static tests used a loaded truck of 21250 kg parked at three bridge positions, while the dynamic tests used loaded and unloaded truck configurations of 21250 kg and 12340 kg at speeds up to 13.9 m/s. The published load-test data are available on Zenodo in HDF5 format and include synchronized measurements from the monitoring system, temporary laser sensors used for vehicle-position and passage-time validation, and derived tracked eigenfrequency time series from the reference and long-term static-load-test period. The bridge geometry, sensor layout, acquisition system, load cases, truck geometry, and data hierarchy are documented to support reuse of the data. Plausibility checks are presented for the main response quantities, including laser-based validation of parking positions and passage windows, vibration spectra and tracked modal parameters, and repeatable strain and displacement response profiles under crawl-speed truck passages. The data set addresses the scarcity of publicly available full-scale bridge load-test data and supports applications such as finite element model calibration, neutral-axis tracking, influence-line extraction, and data-driven anomaly detection.

stat.AP↗

arXiv2024

Confounder-adjusted Covariances of System Outputs and Applications to Structural Health Monitoring

Automated damage detection is an integral component of each structural health monitoring (SHM) system. Typically, measurements from various sensors are collected and reduced to damage-sensitive features, and diagnostic values are generated by statistically evaluating the features. Since changes in data do not only result from damage, it is necessary to determine the confounding factors (environmental or operational variables) and to remove their effects from the measurements or features. Many existing methods for correcting confounding effects are based on different types of mean regression. This neglects potential changes in higher-order statistical moments, but in particular, the output covariances are essential for generating reliable diagnostics for damage detection. This article presents an approach to explicitly quantify the changes in the covariance, using conditional covariance matrices based on a non-parametric, kernel-based estimator. The method is applied to the Munich Test Bridge and the KW51 Railway Bridge in Leuven, covering both raw sensor measurements (acceleration, strain, inclination) and extracted damage-sensitive features (natural frequencies). The results show that covariances between different vibration or inclination sensors can significantly change due to temperature changes, and the same is true for natural frequencies. To highlight the advantages, it is explained how conditional covariances can be combined with standard approaches for damage detection, such as the Mahalanobis distance and principal component analysis. As a result, more reliable diagnostic values can be generated with fewer false alarms.

stat.AP↗

arXiv2024

Nonparametric Monitoring of Spatial Dependence

In process monitoring, it is common for measurements to be taken regularly or randomly from different spatial locations in two or three dimensions. While there are nonparametric methods for process monitoring with such spatial data to detect changes in the mean, there is a gap in the literature for nonparametric control charting methods developed to monitor spatial dependence. This study considers streams of regular, rectangular data sets using spatial ordinal patterns (SOPs) as a nonparametric method to detect spatial dependencies. We propose novel SOP control charts, which are distribution-free and do not require prior Phase-I analysis. To uncover higher-order dependencies, we develop a new class of statistics that combines SOPs with the Box-Pierce approach. An extensive simulation study demonstrates the superiority and effectiveness of our proposed charts over traditional parametric approaches, particularly when the spatial dependence is nonlinear or bilateral or when the spatial data contains outliers. The proposed SOP control charts are illustrated using real-world datasets to detect (i) heavy rainfall in Germany, (ii) war-related fires in (eastern) Ukraine, and (iii) manufacturing defects in textile production. This wide range of applications and findings demonstrates the broad utility of the proposed nonparametric control charts. In addition, all methods in this study are provided as a publicly available \texttt{Julia} package on \href{https://github.com/AdaemmerP/OrdinalPatterns.jl}{GitHub} for further implementations.

stat.ME↗

arXiv2024

Covariate-Adjusted Functional Data Analysis for Structural Health Monitoring

Structural Health Monitoring (SHM) is increasingly applied in civil engineering. One of its primary purposes is detecting and assessing changes in structure conditions to increase safety and reduce potential maintenance downtime. Recent advancements, especially in sensor technology, facilitate data measurements, collection, and process automation, leading to large data streams. We propose a function-on-function regression framework for (nonlinear) modeling the sensor data and adjusting for covariate-induced variation. Our approach is particularly suited for long-term monitoring when several months or years of training data are available. It combines highly flexible yet interpretable semi-parametric modeling with functional principal component analysis and uses the corresponding out-of-sample Phase-II scores for monitoring. The method proposed can also be described as a combination of an ``input-output'' and an ``output-only'' method.

stat.AP↗

arXiv2024

Structural Health Monitoring with Functional Data: Two Case Studies

Structural Health Monitoring (SHM) is increasingly used in civil engineering. One of its main purposes is to detect and assess changes in infrastructure conditions to reduce possible maintenance downtime and increase safety. Ideally, this process should be automated and implemented in real-time. Recent advances in sensor technology facilitate data collection and process automation, resulting in massive data streams. Functional data analysis (FDA) can be used to model and aggregate the data obtained transparently and interpretably. In two real-world case studies of bridges in Germany and Belgium, this paper demonstrates how a function-on-function regression approach, combined with profile monitoring, can be applied to SHM data to adjust sensor/system outputs for environmental-induced variation and detect changes in construction. Specifically, we consider the R package \texttt{funcharts} and discuss some challenges when using this software on real-world SHM data. For instance, we show that pre-smoothing of the data can improve and extend its usability.

stat.AP↗

arXiv2024

Artificial Intelligence in Industry 4.0: A Review of Integration Challenges for Industrial Systems

In Industry 4.0, Cyber-Physical Systems (CPS) generate vast data sets that can be leveraged by Artificial Intelligence (AI) for applications including predictive maintenance and production planning. However, despite the demonstrated potential of AI, its widespread adoption in sectors like manufacturing remains limited. Our comprehensive review of recent literature, including standards and reports, pinpoints key challenges: system integration, data-related issues, managing workforce-related concerns and ensuring trustworthy AI. A quantitative analysis highlights particular challenges and topics that are important for practitioners but still need to be sufficiently investigated by academics. The paper briefly discusses existing solutions to these challenges and proposes avenues for future research. We hope that this survey serves as a resource for practitioners evaluating the cost-benefit implications of AI in CPS and for researchers aiming to address these urgent challenges.

cs.AI↗

arXiv2019

Monitoring data quality for telehealth systems in the presence of missing data

Background: All-in-one station-based health monitoring devices are implemented in elder homes in Hong Kong to support the monitoring of vital signs of the elderly. During a pilot study, it was discovered that the systolic blood pressure was incorrectly measured during multiple weeks. A real-time solution was needed to identify future data quality issues as soon as possible. Methods: Control charts are an effective tool for real-time monitoring and signaling issues (changes) in data. In this study, as in other healthcare applications, many observations are missing. Few methods are available for monitoring data with missing observations. A data quality monitoring method is developed to signal issues with the accuracy of the collected data quickly. This method has the ability to deal with missing observations. A Hotelling's T-squared control chart is selected as the basis for our proposed method. Findings: The proposed method is retrospectively validated on a case study with a known measurement error in the systolic blood pressure measurements. The method is able to adequately detect this data quality problem. The proposed method was integrated into a personalized telehealth monitoring system and prospectively implemented in a second case study. It was found that the proposed scheme supports the control of data quality. Conclusions: Data quality is an important issue and control charts are useful for real-time monitoring of data quality. However, these charts must be adjusted to account for missing data that often occur in healthcare context.

stat.AP↗