Many existing methods for vibration-based monitoring of bridges rely on expensive sensor networks or require direct physical contact with the structure, which limits their widespread adoption, especially across large networks of ageing infrastructures. Recent advancements in non-contact radar technology enable precise and multichannel vibration measurements from a distance, allowing operation safely without interfering with traffic or requiring complex installation. This feature offers a practical and scalable approach to monitoring hard-to-access bridges or structures. However, commercial solutions remain expensive and are often constrained by limited beam coverage, focusing on single-point measurements.

Addressing the existing drawbacks and augmented past research, this study presents an application of low-cost frequency modulated continuous wave (FMCW) radar-based vibration monitoring bridges through experimental validation. This research also investigates multi-point vibration measurements, thanks to the use of multiple transmitting and receiving antennas in radar enabling full-field vibration measurements of the structure. The fundamental working principle of the radar systems is by emitting electromagnetic waves continuously with a linearly changing frequency and comparing its frequency to the reflected waves from the target. To monitor sub-millimetre-level displacements, which are more often observed in the real-world infrastructure systems, the radar systems use the interferometry technique to analyse the phase shift in the reflected signal to accurately estimate the small displacements.

The present study adopts the two-step strategy to validate the predictive capabilities of the FMCW radar system, as shown in Figure. The initial step involves the validation of low-cost sensors against commercial solutions on a corner reflector attached to a shake table, enhancing the signal-to-noise ratio and thereby avoiding interference from static clutter. Subsequently, validation of multiple-point vibration measurements for different targets located within the same range is carried out using multiple transmitting and receiving antennas. This approach increases confidence in the radar system's capability to evaluate the full-field vibration of a structure, rather than relying solely on single-point measurements. Second, the radars are tested for the real-world application on a bridge under the traffic or railway loadings. The field campaign is carried out to evaluate the dynamic behaviour of the bridge, such as its modal properties under ambient traffic conditions. The results reveal that the low-cost radar system effectively captures the dynamic behaviour of both the corner reflectors in the laboratory setup and the bridge during the field campaign. This enables informed decision-making for the stakeholder and asset managers for timely, actionable insight for maintenance and safety interventions by relying on the low-cost FMCW radar systems.

Levees are crucial infrastructures that protect lives and property from flooding. The increasing frequency and severity of extreme flooding events due to climate change have caused the loading scenarios that may exceed the design consideration, leading to a greater risk of distress and malfunction induced, among other factors, by seepage. Variety of monitoring systems have been developed to evaluate the water content in levee material directly, by monitoring, or indirectly by monitoring its temperature. However, existing technologies either require complex installation and maintenance or provide only discrete measurements, which are not suitable for large-scale spatial monitoring. Recent research on RF sensing systems has demonstrated the potential for large-scale spatial water content monitoring. The conceptual vision of implementing RF sensing system for levee humidity monitoring is shown in Figure 1.

This study aims to advance the application of RF sensing system in levee water content monitoring by investigating the relationship between the phase measurement of RF sensing systems and the change of water content within levee’s materials. Previous studies have demonstrated the RF sensing system’s ability to detect the presence of water penetration in dry sand. This study conducts controlled laboratory experiments to quantify the relationship between RF sensors’ phase change and the change in water content in levee material, with particular focus on capturing the nonlinear behavior of the phase change of RF sensors as the water infiltrates the material. The main challenge of this research includes experimental design and analysis of the relationship between the change in water content and the nonlinear behavior of phase change.

The results demonstrate the proof of concept on using passive RF spatial sensing system in levee water content monitoring. The preliminary result shows the relationship between the sensors’ phase measurement and the change in water content. Additional study on the system’s robustness under various noise sources are needed in order to implement the system in the real-world applications.

The demand for renewable energy has grown over the past decade and is projected to accelerate further. This rapid growth has led to an increase in offshore wind deployment and to the installation of increasingly large steel monopiles to account for fatigue damage during installation. Conventional contact-based sensors, such as strain gauges and accelerometers, are often not feasible offshore due to harsh installation conditions and the risk of sensor damage. To address this challenge, a novel non-contact sensor system has been developed that uses magnetomechanical effects to infer strain and an optical sensor to measure velocity. Recent full-scale hammer tests showed that this system can successfully capture strain and velocity during impact piling, allowing for comparison with the industry-standard Pile Driving Analyzer. However, accurate interpretation of magnetometer signals requires understanding the influence of various magnetic disturbances. This work investigates key interference sources arising during offshore pile driving, including the magnetic stray fields of adjacent ferromagnetic components, specifically the outer pile and the hammer-sleeve assembly, as well as the effects of the gradual rotation of the latter. Methods to separate strain-induced magnetization from these other magnetic contributions are explored to enhance signal fidelity. This is achieved through a combination of theoretical calculations and experimental measurements. By examining how these interferences influence strain measurements, this study enhances measurement reliability and advances non-contact monitoring technologies for offshore monopile installations.

Proper appraisal of used automobiles depends on identifying accident and repair histories, yet visual inspections often fail to detect. This study presents an automobile body repair type identification and classification system that combines multi-frequency thermography with TSMixer machine learning. A frequency-modulated laser (2, 5, 6, and 15Hz) applies heat to the automobile body panels, and the resulting thermal response is measured by a high-speed infrared camera. The acquired multi-frequency thermal responses are processed by a TSMixer-based machine-learning model that learns cross-frequency correlations to estimate coating thickness and classify body repair types. The uniqueness of this study are as follows: (1) Development of a noncontact, non-destructive automobile body repair type identification and classification system; (2) Improvement of coating thickness estimation using multi-frequency thermography exploiting frequency–penetration-depth relationships; (3) Development of a robust coating thickness estimation algorithm considering variations in coating material’s properties, automakers, colors, and substrates. Using specimens retrieved from actual used automobiles, the proposed system achieved an accuracy of 0.93 and an F1-score of 0.93 for repair-type classification, alongside coating-thickness estimation with an MAE of 7.31 μm and a MAPE of 4.39 %. These results demonstrate objective, data-driven repair-history inspection, helping customers avoid automobiles with unreported accidents and enabling commercialization of objective value assessment.

Lattice-sandwich structures are widely deployed in aerospace, automotive, and civil engineering owing to their low weight, high stiffness, and multifunctional potential. Nevertheless, the presence of enclosed core architecture presents considerable challenges for conventional vibration-based structural health monitoring techniques. Although the well-developed pseudo-excitation (PE) method, which relies on local dynamic equilibrium, is inherently well suited to complex structural configurations, its accuracy in sandwich structures is compromised by repetitive disturbances originating from face-sheet-to-core joints, as these interferences mask the underlying effective damage index. To overcome this limitation, the present study introduces an improved PE framework that systematically examines the local vibrational responses, characterizes the spectral signatures attributable to joint-related interferences, and eliminates such disturbances through Fourier filtering. Finite-element simulations confirm that the proposed method substantially improves the effectiveness of the PE method in localizing damage within lattice-sandwich structural components. Experimental verification is further conducted using a laser doppler vibrometer. The proposed method preserves the fundamental advantages of the pseudo-excitation approach and broadens its applicability to complex sandwich structures, thereby enabling its practical implementation in engineering applications.

Displacement monitoring represents an important component of structural control and preservation of cultural heritage assets, especially when foundation settlements are involved. The continuous acquisition of displacement data, when integrated with environmental information, allows for the distinction of deformation components associated with environmental variations and those related to structural modifications.

Traditional structural monitoring systems generally rely on the installation of contact sensors directly on the building surface. While this approach is well established, and can provide high-accuracy measurements, its efficacy largely depends on the number, spatial distribution, and proper calibration of the sensors. However, deploying and maintaining a dense network of contact instruments can be costly and logistically demanding. In addition, many historical and artistic monuments are subject to strict conservation regulations that limit or prohibit the installation of invasive instrumentation, thus promoting the development of non-contact monitoring solutions capable of ensuring reliable measurements without altering the physical or aesthetic integrity of the structure. Although non-contact techniques are already mature for large-deformation applications (e.g., slopes and infrastructure), their use in heritage contexts, typically characterized by small displacements and complex geometries, still requires further consolidation and methodological refinement to achieve comparable accuracy and long-term reliability.

Within this framework, the present study investigates the application of Ground-Based Interferometric Real Aperture Radar (GBInRAR) technology as a non-invasive method for the structural monitoring of masonry towers. This technique, developed and refined over the past few decades, enables displacement measurements with sub-millimetric precision by comparing the phase differences of radar signals acquired at successive time intervals. The radar system, positioned at a distance from the target, allows for continuous observation of multiple points along the line of sight, offering a global view of the structural response and temporal evolution of the monitored asset.

The bell tower of the Cathedral of Santa Maria Assunta in Pisa, universally known as the Leaning Tower, was selected as the case study. Over the centuries, the monument has experienced significant settlements and deformations, prompting extensive stabilization works and the establishment of a long-term contact-based monitoring system to ensure its safety. Despite these efforts, several aspects of its structural behavior, and their coupling with environmental variables, remain insufficiently characterized.

Motivated by these gaps, the study presents the results of acquisition campaigns conducted with GBInRAR technology during different periods of the year, along with a comparative analysis of these data and those obtained from the existing monitoring system, while also exploring correlations between observed displacements and environmental parameters. The results aim to highlight the main advantages, limitations, and potential applications of radar-based monitoring in cultural heritage conservation, supporting the development of reliable and non-invasive strategies for the long-term protection and structural behavior assessment of historical monuments.