Radar technology in the microwave and millimeter-wave frequency range is the subject of current research for structural health monitoring of composite materials, e.g., damage detection in wind turbine blades. Performance assessment, enabling widespread practical application of this promising and non-contact sensing approach, can be realized via probability of detection (POD) theory, which is a statistical method for determining the detectability of damage through response metrics as a function of flaw size. This paper deals with the experimental investigation of a delamination model represented by two parallel glass fiber reinforced polymer plates separated from each other from 0mm to 1mm in steps of 0.01mm. Experimental studies with a frequency modulated continuous wave radar are performed under laboratory conditions in the frequency range from 57GHz to 65GHz. The signal response is represented by two damage indicators (DIs), according to the root mean square deviation and Mahalanobis distance. Since the reflection of electromagnetic waves exhibits a nonlinear behavior, this also implies a nonlinear response in the DI characteristic. The novelties in this work are the successful implementation of a nonlinear regression model, combined with an optimal threshold decision through receiver operating characteristic curves for a high-resolution POD representation. The POD with 95% confidence bounds indicates the flaw size at which the delamination can be detected reliably. Depending on the radar distance in experimental studies, the binary structural condition (damaged or undamaged) was correctly assessed from 95% to 100%. The minimum detectable size ranges from 0.01mm to 0.08mm.

This paper evaluates a data-driven classification approach of operational wind turbine blades based on consecutive tower-radar measurements that are each compressed in a two-dimensional slow-time to range representation (radargram). Like many real-world machine learning systems, installed tower-radar systems face some key challenges: (i) transferability to new operational contexts, (ii) impediments due to evolving environmental and operational conditions (EOCs), and (iii) limited explainability of their deep neural decisions. These challenges are addressed here with a set of structured machine learning studies. The unique field data comes from a sensor box equipped with a frequency-modulated continuous wave (FMCW) radar (33.4-36 GHz frequency range). Relevant parts of the radargram that contribute to a decision of the used convolutional neural networks were identified by a class-sensitive visualization technique named GuidedGradCAM (Guided Gradient-weighted Class Activation Mapping). The following main contributions are provided to the field of tower-radar monitoring (TRM) in the context of wind energy applications: (i) every individual rotor blade holds a number of characteristic structural features revealed by the radar sensor, which can be used to discriminate rotor blades from the same turbine via neural networks; (ii) those unique features are not agnostic to changing EOCs; and (iii) pixel-level distortions reveal the necessity of low-level information for a precise rotor blade classification.

We are concerned with the reconstruction of inclusions in elastic bodies based on measurements from a laboratory experiment. In doing so, we solve the inverse problem of the time-harmonic elastic wave equation, in contrast to the stationary wave equation and the corresponding lab experiment proposed earlier in Eberle and Moll (2021). The investigation of the harmonic problem leads to a better reconstruction compared to the stationary one. Since we deal with real measurement data, we have to take into account, that those measurements always include measurement errors, so that we have to handle noisy data. Thus, we consider the linearized monotonicity method for noisy data and introduce a modified version of this method. Based on this, we reconstruct the inclusions numerically.

Structural health monitoring (SHM) using ultrasonic guided waves (UGWs) typically relies on dense sensor networks to enable reliable damage detection and localization, leading to increased system complexity and cost. To address these limitations, this study investigates the use of a frequency-steerable acoustic transducer (FSAT) as a compact alternative capable of directional wave excitation using a single actuator. An experimental investigation is conducted on a conical steel structure representative of industrial geometries. A unidirectional FSAT is employed in a pitch-catch configuration together with a limited number of piezoelectric sensors distributed around the structure. The system is evaluated under pristine and damaged conditions, where a small blind hole is introduced as artificial defect. Guided wave responses are analyzed in both time and frequency domains over a broad frequency range (20–600~kHz), exploiting the FSAT’s frequency-dependent beam steering capability. Damage detection is achieved through differential signal analysis and root-mean-square (RMS)-based damage indicators, which reveal clear sensitivity to the presence and location of the defect. Results demonstrate pronounced frequency-dependent behavior, with a distinct peak response at 350~kHz and significantly reduced sensitivity when the wavefield is steered away from the damage location, confirming directional selectivity. The findings highlight the potential of FSAT-based SHM systems to reduce sensor network requirements while maintaining reliable damage detection and directional sensitivity in complex structural geometries. This work provides an experimental validation of FSAT technology in a realistic structural configuration and represents a step toward simplified and scalable SHM solutions.

Elastic wave control underpins numerous technologies, from structural health monitoring and biomedical imaging to wireless communications and energy harvesting. Conventional phased arrays enable dynamic beamforming but are bulky, power‐intensive, and complex, limiting their integration into compact or distributed systems. Here, we present a meta‐transducer design that embeds wavenumber‐domain filtering directly into the electrode geometry, enabling frequency‐controlled unidirectional generation of ultrasonic guided waves without active phasing networks. A key innovation is the use of error‐diffused spatial dithering to approximate continuous 2D filters with binary patterns by shaping the transducer electrodes, suppressing sidelobes and enhancing angular resolution. Finite element simulations and experimental validation with a scanning laser Doppler vibrometer confirm that the Frequency‐Steerable Acoustic Transducer (FSAT) achieves sharp, frequency‐tunable beam steering across a 180° sector with over 80% reduction in sidelobe energy compared with conventional designs. This approach enables ultra‐compact, energy‐efficient ultrasonic devices with wavefront control, offering a pathway to advanced applications in guided wave imaging, IoT‐enabled sensing networks, and ultrasonic communications.

Environmental and operational conditions (EOC) exert an adverse influence on monitoring systems that utilize permanently installed sensors. This proves particularly evident in the era of machine learning (ML). Since their effects can cause stronger signal changes over time than the phenomena to be detected, EOC must be adequately addressed. Emerging computer-vision (CV) use cases, such as intelligent radar units bound to the mast of wind power plants, share similar challenges with established commercial applications like static video surveillance. Those scenarios suffer from baseline observations for training that may turn out unrepresentative of deployment where such stationary camera-type devices experience temporal data shift due to their evolving non-stationary surroundings. Here, ML-based decisions become prone to false positives, resulting in high downstream investigation costs as well as low user acceptance of these smart systems. Accordingly, a rich literature exists on adapting to various types of drift. However, scenarios remain commonly neglected where (a) only noisy and potentially misleading labels are available as feedback for re-training during operation, (b) no complementary sensors are installed, and (c) changes occur not necessarily in the more obvious marginal distribution of variables, but in their joint probabilities that subtly change over time. This paper therefore focusses on solution strategies involving (i) powerful deep pre-trained models promising robustness (ii) preventive augmentation of training data, (iii) continual learning that constrains model weights over time to mitigate improper shortcut learning. Triggered by recent observations in tower-radar CV research, data drift is studied across several datasets which thereby highlights the problem’s generality. A computational framework is introduced to systematically investigate the complex setup empirically. Probing different ML models, a phase-transition phenomenon in an overfitting tree classifier is incidentally discovered. The paper presents both practical recommendations for CV-based tower-radar monitoring as well as findings of interest for the broader community of researchers in ML.

Wind turbines are critical infrastructure whose economical deployment benefits from blade monitoring. Tower-radar remote sensing generates radargrams from a mast-bound active sensor for this and related purposes. Initial machine learning classifiers are publicly available, yet they have not been stress-tested. Suitable test imagery remains scarce. This paper addresses these gaps through interconnected investigations, supported by two data contributions: a synthetic surrogate benchmark (spanning eight image dimensions in a full-factorial design) and an enrichment of the empirical WiRoRa dataset augmented with human annotations and machine-generated ones where the latter can also serve as an on-the-fly labeling solution in the field. The studies report: multi-class anomaly filtering; a sensitivity analysis revealing that the end-to-end-trained (E2E) classifier is poorly calibrated, while its pretrained counterpart is substantially more stable; adversarial vulnerability evaluation showing that the E2E model is also more easily fooled; an analytical derivation of when/why bottleneck training on auxiliary imagery improves representations; a surrogate test confirming the bottleneck hypothesis; and a preliminary Mixture-of-Experts pilot for enhanced traceability as well as scalability that performs environmental/operational metaparameter regression as an archetypal example. Together, the results expose failure modes of existing classifiers and chart a path toward intrinsically interpretable systems for structural health monitoring and beyond.

This work presents an experimental investigation of a stand alone directional structural health monitoring (SHM) system based on a frequency-steerable acoustic transducer (FSAT) operating in pulse-echo mode. A single transducer bonded at the center of a 1000 & times; 1000 & times; 1 mm aluminum plate was used to perform damage detection and localization by exploiting the frequency-dependent beam steering characteristics of the FSAT. Directional scanning of the surrounding structure was achieved without mechanical movement or phased-array electronics. Damage detection employed baseline comparison using a root mean square metric applied to reflected Lamb wave signals. Both directional detection at fixed actuation frequencies and spatial scanning via frequency sweeps were experimentally demonstrated. The system successfully detected and localized multiple damage types, including drilled holes and attached disk masses with a diameter of mm, with characteristic dimensions as small as 3 mm. Progressive detection and discrimination of single, double, and triple damage scenarios at distinct angular locations validated the FSAT's capability for complex multidamage monitoring. The localization accuracy was consistent with the transducer's angle-frequency calibration, and time-of-flight measurements provided the radial distance estimation capabilities. Compared to conventional omnidirectional sensor networks or phased-array systems, the presented solution offers reduced system complexity, lower sensor density, and improved scalability, while remaining compatible with existing ultrasonic sensor networks. These results demonstrate the potential of the FSAT for simpler, lighter, and more scalable autonomous SHM systems, with enhanced inspection flexibility and localization accuracy suitable for industrial applications.

Composite overwrapped pressure vessels are increasingly being used for hydrogen storage. However, durability and safety assessment are crucial during the lifetime. In this view, the ultrasonic guided wave (UGW)-based structural health monitoring is applied to a pressure vessel during burst testing campaign to investigate the evolution and behavior of guided wave propagation under extreme loading conditions up to final failure. A full-scale vessel is instrumented with a distributed network of piezoelectric transducers, and guided wave measurements are conducted throughout a controlled pressurization cycle until burst. Signal changes are evaluated as a function of the pressurization cycle and progressive structural degradation. Guided wave features processed using machine learning algorithms, together with selected damage indices, are analyzed to assess their sensitivity and behavior under near-burst conditions. The results show the sensitivity of guided waves to extreme pressure conditions. In addition, the study highlights the challenges due to multi-modal behavior of Lamb waves, which needs to be accounted for in metrics selection. Valuable support in visualizing structural degradation under extreme load is provided by principal component analysis, which allows clustering data and, as such, highlighting structural changes. This work provides valuable experimental reference for the validation of guided wave-based Structural Health Monitoring strategies for Composite overwrapped pressure vessels under critical conditions.

Tower-radar computer vision (TRCV) represents an emerging application-oriented field of study. Here, image-type measurements are acquired from radar transceivers bound to the mast of wind power turbines and these radargrams subsequently get analyzed using data-driven algorithms. TRCV shows promise for reducing downtime and increasing safety of such renewable energy plants, accordingly, the respective monitoring systems should long-term perform real-time recognition of moving objects like rotor blades and their condition, disregard maintenance workers, identify birds and bats or unauthorized aerial vehicles, in order to trigger appropriate multi-faceted responses. This article focuses on the radar-only computer-vision task of classifying rotors without complementary costly instrumentation [1]. Progress in TRCV for blade monitoring remains, however, hindered by the limited publicly available data [2]. For this specialized task, recently [3] general-purpose feature extractors have proven robust to environmental and operational conditions and as a valuable building block in TRCV processing pipelines. Such large models, like OpenCLIP, have been pre-trained on internet-scale amounts of text-image pairs. They however (i) commonly still require additional data for transfer to dedicated use cases, as well as (ii) exhibit power demands or latencies incompatible with real-time resource-constrained application scenarios, models hence need to be compressed. In this paper, measured radargrams from field experiments are therefore complemented with the novel synthetic dataset SiWiRoRa as well as further open imagery. Here, several specialized image-type datasets are carefully compiled to span a context around the focal measured radargrams. Second, the main geometrical directions of this context landscape are explained by classical image statistics and with human perceptions collected from annotators. Third, large models are distilled towards lightweight capacity where it is found that both the synthetic dataset but also seemingly unrelated imagery can improve performance in blade classification. Accordingly, routes for enhancing the SiWiRoRa dataset are suggested and moreover implemented so as to further advance TRCV. References [1] Alipek, Sercan, et al. "Potential and Limitations of Anomaly Detection via Tower-Radar Monitoring of Wind Turbine Blades in Regular Operation with Convolutional Networks." EWSHM (2024). [2] Mälzer, Moritz et al. "Radar-based structural monitoring of wind turbines blades: Field results from two operational wind turbines." IWSHM (2023). [3] Kexel, Christian et al. "Mast-Bound and Too Curious: Overcoming Drift in Wind-Tower Radar for Blade Monitoring Using Pre-Training, Augmentation and Weight Consolidation Due to Correlated Conditions." LATAM-SHM (2026)

Ultra-wideband electromagnetic waves are used in many applications of structural health monitoring (SHM). The inspection of large glass fiber composite structures poses a significant challenge and requires additional research and development to guarantee a safe and reliable operation. SHM based on a permanent sensor installation and combined with advanced data analysis approaches can help to ensure structural integrity. Therefore, a widely used approach is to evaluate the structural condition by comparing measurements from the healthy state with the current state of the structure. In this work, a novel reversible reference damage model for a delamination is presented that can be detected with a frequency modulated continuous wave (FMCW) radar at 60 GHz. The reversible scatterer is a sandwich model made of Rohacell((R)) material and erosion protection tape specially developed for field application in wind turbine blades (WTB). The reference damage model was successfully tested numerically and experimentally in preliminary investigations. A damage indicator (DI), represented by the root mean square deviation (RMSD) of two structural states, i.e. healthy and damaged, is calculated to detect a reference damage in a WTB segment.

This paper describes a system for detecting damage in the rotor blades of wind turbines. It combines two sensor technologies for this purpose: millimeter wave radar detects local structural damage in the walls of the rotor blades, while an acceleration sensor identifies global deviations in vibration behavior, for example in the case of ice on the blades. A complete description of the measurement system and measurement environment is provided, with a particular focus on the electronic architecture of the sensor nodes, based on the power over dataline standard.

Delivery drones have become increasingly important in recent years. It is advantageous for commercialization that suppliers are able to deliver orders autonomously and directly to their customers via air transport. However, the safety aspect must be considered. Real-time inspection of delivery drones during operation helps preventing accidents and a threat to civilians. For continuous monitoring, the sensors must be installed on the drone throughout the flight. A promising approach uses the inherent excitation of the servomotors for vibration-based Structural Health Monitoring (SHM). Vibrations can be recorded using triaxial acceleration sensors and analyzed using suitable methods such as stochastic subspace-based fault detection or histogram difference. In comparison to nondestructive testing, reference measurements of the intact structure are necessary in SHM. As soon as laboratory conditions no longer exist and environmental parameters, for example, wind, influence the vibration spectrum, classification methods are necessary for compensation. The recording of comprehensive reference datasets with different environmental conditions is limited by the battery life. This work focuses on diagnosing irreversible rotor blade damages of an 8.5 kg delivery drone. A parametric analysis taking into account systematic fusion of damage indicators and a specific number of considered references were determined for this purpose in order to assess the severity of the existing damage. Onboard SHM to evaluate the airworthiness in real time for linear and hovering flights was achieved.

This work leverages ultrasonic guided waves (UGWs) to detect and localize damage in structures using lightweight Artificial Intelligence (AI) models. It investigates the use of machine learning (ML) to train the effects of the damage on UGWs to the model. To reduce the number of trainable parameters, a physical signal processing approach is applied to the raw data before passing the data to the model. Starting from current state of the art in algorithms used for damage detection and localization, an AI-based technique is developed and validated on an experimental benchmark dataset before tiny ML implementation on a low-cost development board. A discussion of the need for a balance between the reduction in computational resources and increasing the precision of the models is also reported. It is shown that by extracting simple features of the signal, the models required to predict the damage locations can be significantly reduced in size while still having high accuracies of over 90%. In addition, it is possible to use these predictions to construct a fairly accurate heat map indicating the likely damage locations. Finally, a convenient edge/cloud visualization of the results can be achieved by simplifying the heat map.

In recent years, the development of machine learning (ML) techniques has led to significant progress in the field of structural health monitoring with ultrasonic-guided waves. However, a number of challenges still need to be resolved for reliable operation in realistic settings. In this work, we consider the complex problem of experimental damage detection under varying temperature or load conditions where damage locations are not included in the training set. The ML techniques proposed here include supervised and unsupervised methods originally developed for image and time series classification combined with ensemble voting. A performance demonstration of the ML techniques is presented using benchmark datasets from the open-guided waves platform. The unsupervised approach is then applied to a new dataset from an experimental campaign carried out on a composite over-wrapped pressure vessel used for hydrogen storage with real defects. Results show that ensemble voting enables the effective combination of the predictions of multiple transducer pairs, even with a limited number of strong individual classifiers. When applied to unsupervised learning, this returns high accuracy also when real damage over the structure is considered.

Ultrasonic guided waves (GWs) are extensively utilized in nondestructive evaluation and structural health monitoring (SHM) fields. Typically, phased-array GW-based inspections consist of numerous piezoelectric transducers permanently attached to the monitored structure. However, these systems face challenges such as bulky hardware, a large number of transducers and cables for individual element control, complex circuitry and signal processing, high power consumption, and consequently high integration costs. To overcome these limitations, shaped transducers featuring inherent beam steering properties, such as Frequency Steerable Acoustic Transducers (FSATs) can be adopted. FSATs exploit a frequency-dependent spatial filtering effect, which is achieved by properly patterning the electrodes of the piezoelectric transducers. This allows the direction of the generated or sensed wave to be controlled simply by the spectral content of the actuated or received signal, a process so-called "In-sensor" signal processing. Initial generations of FSATs face a 180 degrees ambiguity, where waves are simultaneously generated or sensed in both forward and backward directions. This could lead to uncertainty in defect localization or generate undesirable reflections. In this work, a novel unidirectional FSAT is proposed to eliminate this ambiguity through a new design strategy for unidirectional wave generation and sensing, addressed in the wavenumber domain. Finite element simulations and experimental testing on an aluminum plate validated the proposed frequency-dependent unidirectional beam steering concept. Additionally, the transducer was successfully used in pulse-echo mode for damage imaging, demonstrating 98% localization accuracy. The proposed embedded system can substantially reduce the software and hardware requirements of conventional solutions, paving the way for the development of permanent inspection systems.

This paper reports a convolutional neural network (CNN)-based damage detection approach for radar-based structural health monitoring of wind turbine blades. Subsequent radar measurements are transformed into an image-type representation for use as CNN input. In contrast to conventional approaches that require compensation for temperature and loading effects, the proposed framework inherently learns all required information during the training phase. Its damage detection performance (i.e., detecting intact vs. damaged condition) is demonstrated using measurements from multiple embedded radar sensors during fatigue testing of a wind turbine blade with a length of 31 m. The achieved F1-score for correct damage classification is between 91% and 100% for both the unloaded and the loaded blade.

Maintaining the structural integrity of composite overwrapped hydrogen pressure vessels (COPVs) is increasingly critical for safe hydrogen storage and transportation, given the unique challenges posed by composite materials. Accurate structural assessment is vital for effective Structural Health Monitoring (SHM), and guided ultrasonic waves (GUW) show significant potential for this application. This study presents a model-assisted approach to damage assessment using guided wave-based SHM (GWSHM) on COPVs. By combining experimental data, numerical simulations, and machine learning techniques, we aim to enhance the reliability and accuracy of detection capabilities. Experiments were conducted on both intact vessels and those with artificial damage to generate baseline and damage-induced guided wave signals. Two simulation techniques were employed to support these measurements. Another key contribution of this work is the integration of model-based insights with data-driven approaches for damage assessment. The findings contribute to the advancement of GWSHM methodologies and support the safe, efficient deployment of COPVs in hydrogen storage applications.

This paper characterizes a system for damage detection on wind turbine rotor blades. The system utilizes a millimeter wave radar for the detection of structural damage in wind turbine blades in combination with an acceleration sensor, which detects changes in vibration behavior, e.g., in the case of ice on the rotor blades. The direction-dependent attenuation behaviour of the radar within the sensor node housing is determined, and the transmission of initial telemetry data from a first prototype of the system, which is installed in an operating wind turbine, is described and successfully tested.

Despite proven approaches available in the literature, structural health monitoring by ultrasonic guided waves under varying environmental and operational conditions is still challenging. The use of machine learning approaches is discussed in this work, considering the complex problem of experimental damage detection under varying load conditions in a composite overwrapped pressure vessel for hydrogen storage. Specifically, unsupervised methods originally developed for image and time series classification are combined with ensemble voting to conceive reliable damage detection technique. This enables the effective combination of the predictions of multiple transducer pairs, even with a limited number of strong individual classifiers. A performance demonstration of the technique is presented using a real damage scenario dataset.