Kalman filter-based residual generation followed by statistical change detection is a standard approach for model-based cyberattack detection. The methods are often tested in idealized simulation studies. In the case of tests in laboratory test beds, typically a detailed analysis of the signal properties is lacking. In this contribution, said approaches are tested in a model factory that uses smart industrial sensors, communication, and control systems such that the signal properties are close to real-world application. It is shown that amplitude quantization effects and temporal correlation occur that affect attack detection.

A dataset was collected to investigate the reproducibility of ultra-high performance concrete (UHPC) quality under controlled variability, since UHPC mechanical properties can vary even when using a fixed mix formulation in production. This dataset is the first open-access dataset in the UHPC community to take a holistic view of the UHPC production process and associated production variability, featuring a relatively large number of systematically designed experiments under a fixed recipe. A reference UHPC recipe, designed for a 28-day compressive strength of 120 MPa, was used to produce 150 experimental records, while the quality, storage conditions, and dosing accuracy of raw materials, along with mixing parameters and curing conditions, were systematically varied to emulate real-world production environments. For each batch, five fresh UHPC properties and the average mixer power were recorded, and compressive and flexural strengths at 24 hours and at 28 days were measured using standardized methods under controlled laboratory conditions. By linking process variability to mechanical outcomes under the fixed reference recipe, the dataset supports reproducibility analysis in UHPC production and provides reusable data for machine learning, uncertainty modeling, and digital-twin development.

The industrial progress of laser based metallic directed energy deposition (DED-LB/M) depends heavily on robust in-situ quality assurance. While thermography offers high spatial and temporal resolution for process monitoring, its transition from a qualitative observation tool to a quantitative measurement system is hindered by a fundamental inversion problem. In laser-based metal deposition, the spectral emissivity evolves dynamically due to rapid phase transformations, extreme temperature gradients, and surface oxidation, leading to significant temperature measurement errors when using conventional single-color pyrometers and thermographic cameras.This study presents a multispectral thermographic approach to decouple temperature and emissivity during the DED-LB/M process. For the analysis of the quasi simultaneously captured thermal radiation across multiple spectral bands, a Temperature Emissivity Separation (TES) algorithm is employed to solve the underdetermined system of equations. The methodology is evaluated using data of the laser-based deposition of 316L, demonstrating that the multispectral approach can effectively compensate for and determine emissivity fluctuations that would otherwise result in temperature deviations of several hundred Kelvin. However, due to its complexity, this method should rather be regarded as reference method for other sensorics or simulation validation than as a monitoring technique.

This paper proposes an ensemble-based uncertainty quantification framework for Gaussian Process Regression in small-data regression problems, where reliable uncertainty estimates are essential. The method enables a systematic decomposition of predictive uncertainty by capturing variability induced by model hyperparameters while retaining the data efficiency of standard GPR. By interpreting the ensemble as a finite mixture of Gaussian Processes, the approach provides a principled and computationally tractable alternative to fully Bayesian inference. The framework is demonstrated on a micromagnetic sensor calibration task, highlighting its ability to deliver robust predictions together with meaningful uncertainty information. The results show that the proposed uncertainty-aware modeling approach improves reliability in data-scarce settings and supports informed decision-making in material property prediction.

Abstract The size-of-source effect (SSE) is a systematic deviation causing infrared temperature measurements to vary with radiant source size. Existing formulations typically focus on diffraction, often omitting scattering contributions. This work proposes a hybrid, data-driven model defined by three parameters estimated via least squares regression, aiming to provide a tool for SSE characterization and compensation in practical thermography. Results indicate incorporating scattering in model formulation significantly improves the description of the SSE. Sensitivity analysis identified distinct parameter dominance ranges, and model robustness within the solution space was confirmed. Moreover, residuals calculated against the prediction remained below the camera’s measurement uncertainty. Finally, validation against a second dataset confirmed the model’s capability to capture primary SSE behavior, thereby demonstrating its potential for SSE compensation in acquired thermograms.

The industrial real-time application of the non-destructive testing method of micromagnetic material characterization offers great potential for the analysis of near-surface workpiece properties. In addition to possible reductions in time and cost through the direct integration of testing methods into the machining process, which so far has only been possible offline, this enables monitoring and targeted control of relevant workpiece properties during the ongoing process. A central challenge in the application of micromagnetic material characterization methods (MMV), also due to small reference datasets, is the development of calibration models. MMV can, for example, be applied in process data acquisition for process parameter optimization in manufacturing, without the need for complex ex-process reference measurements. In this contribution, a data-driven approach to calibration model development for micromagnetic material characterization is presented, in which Gaussian Process Regression (GPR) is combined with various feature selection methods. The proposed methodology is evaluated using an industrial hard turning process and assessed in terms of its performance compared to classical approaches such as Multiple Linear Regression (MLR).

Ultra-high performance concrete (UHPC) surpasses conventional concrete in performance. However, ensuring consistent mechanical properties during production, even with identical recipes, remains challenging. Using experimental data, this study investigates how material quality, environmental conditions, measurement errors in material dosing, and mixing and curing conditions influence the mechanical properties of UHPC. This broad scope of influencing factors and production conditions increases data dimensionality and, coupled with the high cost of UHPC experiments, results in a sparse dataset. Traditional evolutionary algorithms, though effective in feature selection, struggle with high-dimensional small-sized datasets. To address this, a search-space-constraining method for the non-dominated sorting genetic algorithm II (NSGA-II) is introduced, incorporating domain-specific knowledge into population initialization to reduce dimensionality and thus enhance prediction accuracy and solution stability. Comparative evaluations using various machine learning algorithms on the UHPC dataset demonstrate that population initialization to constrain the search space of NSGA-II outperforms the standard NSGA-II. Finally, the significance of each examined factor in the UHPC manufacturing process for the properties of the final product is discussed. Ultra-Hochleistungsbeton (UHPC) & uuml;bertrifft konventionelle Betone hinsichtlich der Leistungsf & auml;higkeit. Die Gew & auml;hrleistung reproduzierbarer mechanischer Eigenschaften bleibt jedoch selbst bei identischen Rezepturen herausfordernd. Auf Basis experimenteller Daten untersucht diese Studie den Einfluss von Materialqualit & auml;t, Umgebungsbedingungen, Messfehlern in der Materialdosierung sowie Misch- und Nachbehandlungsbedingungen auf die mechanischen Eigenschaften von UHPC. Die Vielzahl potenzieller Einflussgr & ouml;ss en und Prozessbedingungen f & uuml;hrt zu hoher Dimensionalit & auml;t; zugleich begrenzen die hohen Versuchskosten die Stichprobengr & ouml;ss e. Es liegt daher ein hochdimensionaler Datensatz mit geringer Fallzahl vor. Klassische evolution & auml;re Algorithmen sind zwar in der Merkmalsauswahl leistungsf & auml;hig, sto ss en bei hochdimensionalen, kleinen Datens & auml;tzen jedoch an Grenzen. Zur Abhilfe wird eine suchraumbeschr & auml;nkende Variante des Non-Dominated Sorting Genetic Algorithm II (NSGA-II) vorgestellt, die dom & auml;nenspezifisches Wissen bereits in der Populationsinitialisierung einbindet, die effektive Dimensionalit & auml;t reduziert und damit Vorhersagegenauigkeit sowie L & ouml;sungsstabilit & auml;t erh & ouml;ht. Vergleichende Auswertungen mit verschiedenen Machine-Learning-Verfahren auf dem UHPC-Datensatz belegen Vorteile gegen & uuml;ber dem Standard-NSGA-II hinsichtlich Vorhersageg & uuml;te und L & ouml;sungsstabilit & auml;t. Abschlie ss end wird die relative Bedeutung der betrachteten Einflussfaktoren im UHPC-Herstellprozess f & uuml;r die Eigenschaften des Endprodukts diskutiert.

Der industrielle Echtzeiteinsatz des zerstörungsfreien Prüfverfahrens der mikromagnetischen Materialcharakterisierung bietet ein großes Potenzial für die Analyse oberflächennaher Werkstückeigenschaften. Neben einer möglichen Zeit- und Kostenreduktion durch die direkte Integration von Prüfverfahren in den Bearbeitungsprozess, die bisher nur offline durchführbar war, ermöglicht dies eine Überwachung und gezielte Regelung relevanter Werkstückeigenschaften im laufenden Prozess. Eine zentrale Herausforderung beim Einsatz mikromagnetischer Materialcharakterisierungsverfahren (MMV) ist dabei auch auf Grund kleiner Referenzdatensätze die Kalibriermodellerstellung. Die MMV können beispielsweise bei der Prozessdatenerfassung für die Prozessparameteroptimierung in der Fertigung eingesetzt werden, ohne dabei aufwendige ex-process Referenzmessungen durchführen zu müssen. In diesem Beitrag wird ein datengetriebener Ansatz zur Kalibriermodellerstellung für die mikromagnetische Materialcharakterisierung vorgestellt, bei dem Gaußprozessregression (GPR) mit verschiedenen Verfahren zur Merkmalsselektion kombiniert wird. Die vorgestellte Methodik wird anhand eines industriellen Hartdrehprozesses evaluiert und im Hinblick auf ihre Leistungsfähigkeit gegenüber klassischen Verfahren wie multipler linearer Regression (MLR) bewertet.

Ultra-high performance concrete (UHPC) combines exceptional strength and durability, yet its industrial production is hampered by batch-to-batch variability that generates costly off-specification waste. Leveraging a 150-batch design-of-experiments dataset based on systematic variations of a single reference UHPC mix, this study takes a holistic view of the UHPC manufacturing chain and quantifies how fluctuations in raw material quality, storage conditions, dosing errors, mixer energy demand, and curing regimes affect the 28-day compressive strength. Ten diverse machine learning algorithms are benchmarked; the best-performing model explains $$\ge$$ 75 % of the strength variance with a prediction error $$\le$$ 10 % under leave-one-out cross-validation. SHapley Additive exPlanations reveal that long-term curing temperature and humidity dominate strength development, followed by ingredient moisture and silica fume impurity. These insights are operationalized in an at-line, operator-in-the-loop recommendation system that explores the curing envelope and proposes end-of-mix, batch-specific adjustments before curing starts. In five validation cases, curing adjustments rescued 5/5 underperforming batches, eliminating 75 L of off-specification UHPC and—considering cement only with 600 kg/$$\mathrm {m^{3}}$$ and 15 L per batch of UHPC made with white Portland cement—avoided $$\approx$$ 41 kg CO2e (cement-only; 0.913 kg CO2e/kg, A1–A3). The framework therefore not only elucidates the main sources of UHPC quality inconsistency but also provides a practical, data-driven tool to rescue off-specification products, minimize waste, and cut associated $$\mathrm {CO_2}$$ emissions.

Current radiometric calibration procedures for thermal imaging cameras are typically performed for a single calibration geometry. Typical measurement scenarios do not replicate this geometrical condition and the values provided by the devices deviate from the calibration, which is known as the size-of-source effect (SSE). This study presents SSE measurements on four thermal imaging cameras: two operating in the long-wavelength infrared (LWIR) range and two in the mid-wavelength infrared (MWIR) range. The results indicate that the SSE increases with temperature as the size of the object deviates from the calibration geometry. The deviations observed in the MWIR cameras were smaller than those in the LWIR cameras. The horizontal modulation transfer function (MTF) was measured and utilised to simulate the response of the optical systems and to compare predictions from an analytical approach with the SSE measurements. Although similar trends are observed between the measured and predicted SSE, the model still fails to reproduce the magnitude of the deviations, possibly due to effects not accounted for by the one-dimensional model, along with additional factors such as scattering. Future work will focus on formulating a 2D model and incorporating additional terms into the system transfer function to account for effects not covered by the measured MTF and to enhance the theoretical prediction of the SSE.

An airborne measurement system with an onboard computer for data processing and recording that does not require constant radio communication for inspection and maintenance is presented. It detects, locates, and quantifies methane leaks using a gimbal-mounted tunable diode laser absorption spectroscopy (TDLAS) sensor. A polynomial regression model that correlates wind speed with drone attitude is presented and compared to measurements made with a 3D anemometer at varying wind speeds. The quantification of methane emissions was evaluated with the system, both in a laboratory setup and at an open-area test site.

One of the concerns in the additive manufacturing of metallic pieces is the lack of quality assurance and consistency. To address this, it is necessary to have closed-loop control of the temperature and cooling rate not only at a single point but throughout the whole workpiece. This paper presents a method to create control-oriented spatio-temporal temperature models for 2D workpieces in Direct Laser Deposition and identify them with thermal imaging data. Data from laboratory experiments during the production of workpieces of two different geometries were used to validate the method. An emissivity compensation method was devised to correct the temperature estimation from the infrared measurement around the melt pool area. The performance of the models obtained was assessed so as to suggest further improvements.

The plastic injection molding process has been established as the most widespread manufacturing process in the plastic processing industry. It is employed by almost 70 % of all plastic processing companies [1], [2]. Among the decisive factors contributing to its prevalence are the ability to manufacture parts with intricate geometries and a high degree of automation. There are approaches of varying complexity to control the part quality to reduce waste and increase the efficiency of the process: The industry standard is the control of the so-called machine-variables, i.e., the process variables that are measured on the machine side of the process. This does not take into account any variables that reflect the true state of the emerging part. For this reason, the scientific community aims to control process variables that are measured cavity-side, more precisely, the pressure in the mold cavity. However, the implementation of pressure control requires significant control knowledge and is not suitable for large-scale industrial application. The objective of this contribution is therefore to transform an ordinary machine-variable controlled injection molding machine to a Cyber Physical Production System (CPPS) via augmentation by a digital twin (DT). The DT will predict the part quality from process variables. To this end, a state-of-the-art industrial injection molding machine was equipped with additional sensors that measure in-cavity process variables. Moreover, an in-line quality measuring cell was added. By doing so all machine, processes, and quality data required for data-driven modeling and perspectively control are acquired. Subsequently, an internal dynamics approach for predicting final batch quality from process value trajectories is proposed and compared to the current state-of-the-art modeling approaches in a case study.

Infrared thermal imaging enables fast, accurate and non-contact measurement of temperature distributions. However, 2D representations of 3D objects often require several images to provide significant information. For such cases, 3D thermograms allow a quick temporal and spatial analysis. In this paper, the integration of an industrial high-precision 3D sensor into a 3D thermography system is presented. The performances of the existing and new systems are assessed and compared by analyzing 3D thermograms of an industry-related test object. The geometry of the obtained point cloud is evaluated by means of a non-referenced point cloud quality assessment approach. It is shown that, in the presence of the spatial resolution and the local curvature, the proposed system performs significantly better than the existing one.

Data-driven approaches are an effective solution for modeling problems in machining. To increase the service life of hard-turned components, it is important to quantify the correlation between the cutting parameters such as feed rate, cutting speed and depth of cut and the near-surface properties. For obtaining high-quality models with small data sets, different data-driven approaches are investigated in this contribution. Additionally, models that enable uncertainty quantification are crucial for effective decision-making and the adjustment of cutting parameters. Therefore, parametric multiple polynomial regression and Takagi–Sugeno models, as well as non-parametric Gaussian process regression as a Bayesian approach are considered and compared regarding their capability to predict residual stress and surface roughness values of 51CrV4 specimens after hard-turning. Moreover, a novel method based on optimization of data driven non-linear models is proposed that allows for identification of cutting parameter combinations, which at the same time lead to satisfactory surface roughness and residual stress states.

Der industrielle Echtzeiteinsatz zerstörungsfreier Prüfverfahren bietet ein großes Potenzial für die Analyse oberflächennaher Werkstückeigenschaften. Neben einer möglichen Zeit- und Kostenreduktion durch die direkte Integration von Prüfverfahren in den Bearbeitungsprozess, die bisher nur offline durchführbar war, ermöglicht dies eine Überwachung und gezielte Regelung von relevanten Werkstückeigenschaften im laufenden Prozess. Eine zentrale Herausforderung beim Einsatz mikromagnetischer Materialcharakterisierungsverfahren (MMV) [1] ist dabei deren Kalibriermodellerstellung. Die MMV können beispielsweise bei der Prozessdatenerfassung für die Prozessparameteroptimierung in der Fertigung eingesetzt werden [2], ohne dabei aufwendige ex-process Referenzmessungen durchführen zu müssen. In diesem Beitrag wird ein Ansatz für die datengetriebene Kalibriermodellerstellung für die mikromagnetische (MM) Messung vorgestellt und anhand eines Hartdrehdatensatzes untersucht.

Concrete is an essential material ubiquitously employed in construction. Yet, deciphering the factors that influence its quality is a formidable challenge due to partially understood physical relationships, the high dimensionality of the data, and its limited availability. This study introduces an ensemble framework designed to address these challenges. It uses a combination of individual methods within an ensemble configuration to identify the critical features that determine concrete quality. Within this framework, diverse base methods are harmonized using an average-based technique, leading to a robust final verdict. After selecting the potential influencing factors, 50 experiments are conducted using the Taguchi Orthogonal Array (L-50) to generate the data points. The proposed ensemble learning framework underscores the substantial impact of storage conditions during the curing time on the final quality of concrete.