(1) Background: The purpose of this review is to explore how advanced sensor technologies and AI-driven methods, like machine learning and image processing, are shaping non-destructive imaging (NDI) systems. NDI plays a vital role in ensuring the strength and reliability of composite materials. Recent advancements in sensor technologies and AI-driven methods, such as machine learning and image processing, have opened up new ways to improve NDI systems, offering exciting opportunities for better performance. (2) Methods: This review takes a close look at how advanced sensor technologies and machine learning techniques are being integrated into NDI systems. The review evaluates how effective these technologies are at detecting defects and examines their strengths, limitations, and challenges. (3) Results: Combining sensor technologies with AI methods has shown a clear boost in defect detection accuracy and efficiency. However, challenges like high computational requirements and integration costs remain. Despite these hurdles, the potential for these technologies to revolutionize NDI systems is significant. (4) Conclusions: By synthesizing the latest research, this review offers a comprehensive understanding of how sensor technologies are enhancing NDI. The findings highlight their importance for improving defect detection and their broader impact on research and industry, while also pointing out areas where further development is needed for future growth.

In order to evaluate the short-term voltage stability of an electrical power grid, it is necessary to employ not only systematic and well-targeted fault simulations, but also an evaluation method that assesses the criticality of the individual scenarios. A binary decision between stable or unstable, or whether a threshold value is exceeded or not, is inadequate, particularly in instances where the modeling of the system is subject to a certain degree of uncertainty. Since systematic deviations are subject to natural principles and an intervention limit can thus be determined deterministically, an evaluation method is therefore required that allows a statement to be made about the proximity to instability or to a threshold value. It is common practice to employ indices for the evaluation of voltage recovery following a fault event in simulations or from real measurements. However, depending on the specific question being analyzed, the requirements for an index may vary. A review of the literature revealed the existence of several indices that have been developed and applied in the context of various problems and analyses. These indices have been shown to be effective in the respective contexts. However, none of them fully satisfy the requisite criteria for addressing the aforementioned issue. This paper presents and discusses a new index that was developed explicitly for the problem at hand in dealing with model uncertainties, derived requirements from it, and compared it with existing indices from the literature. The benefits of this novel index in comparison with the established ones were visualized based on a number of indicative simulations. Subsequently, the uncertainties inherent in the load parameterization and their implications on the voltage recovery were presented via Monte Carlo simulations. The evaluations of these effects in terms of the distance from the permissible threshold value were then analyzed using the various indices. All simulations were executed within the framework of the IEEE 39 bus New England system.

This work proposes a nonlinear modeling of a permanent magnet synchronous motor (PMSM) based on state space neural networks. The state space neural network is trained and the state variables (currents in a direct–quadrature frame and the rotational speed) are estimated by considering a robust Unscented Kalman Filter (UKF). Two contributions are presented in this work: the fist one is a nonlinear modeling structure for a PMSM based on a state space neural network that allows real-time parameter identification, and the second one is PMSM neural network training and state estimation based on a robust UKF. The robustness of the UKF is obtained by using a singular value decomposition of the covariance matrix. A comparison analysis is performed over a real PMSM motor by considering the proposed approach and a linear approximation of the nonlinear model where the states and parameters are computed by using an Extended Kalman Filter. The identified model is validated in closed loop by considering a nonlinear control strategy based on state feedback linearization.

The development of urban areas has led to an increase in the use of subsoil for installing transportation networks. These systems usually comprise the construction of side-by-side twin running tunnels built sequentially and in close proximity. Different studies have demonstrated that under such conditions, there is an interaction between tunnels, leading to greater settlements compared with those obtained if the tunnels were excavated separately. Supported by those findings, several analytical methods have been proposed to predict the settlements induced by the excavation of the second tunnel. This paper examines the applicability of these proposals across multiple case studies published in the literature by comparing the analytical predictions with the reported monitoring data of 57 sections. The results indicate that, regardless of the different soil conditions and geometrical characteristics of the tunnels, a Gaussian curve accurately describes the settlements in greenfield conditions and those induced by the second tunnel excavation, although with the curve becoming eccentric in this case. Despite some significant scatter observed, most methods predict the settlements induced by the second tunnel with reasonable accuracy, with Hunt’s method presenting the best fit metrics. The obtained findings confirm that existent methods can be a valid tool to predict the settlements induced by twin tunnelling during the early stages of design, although do also contain limitations and pitfalls that are identified and discussed throughout the paper.

The aim of this Eng Special Issue is to collect experimental and theoretical research relating engineering science and technology to the general topics of Eng [...]

This study evaluates the effectiveness of advanced statistical and geospatial methods for analyzing radon concentration distributions in indoor environments, using the district of San Martín de Porres, Lima, Peru, as a case study. Radon levels were monitored using LR-115 nuclear track detectors over three distinct measurement periods between 2015 and 2016, with 86 households participating. Detectors were randomly placed in various rooms within each household. Normality tests (Shapiro–Wilk, Anderson–Darling, and Kolmogorov–Smirnov) were applied to assess the fit of radon concentrations to a log-normal distribution. Additionally, analysis of variance (ANOVA) was used to evaluate the influence of environmental and structural factors on radon variability. Non-normally distributed data were normalized using a Box–Cox transformation to improve statistical assumptions, enabling subsequent geostatistical analyses. Geospatial interpolation methods, specifically Inverse Distance Weighting (IDW) and Kriging, were employed to map radon concentrations. The results revealed significant temporal variability in radon concentrations, with geometric means of 146.4 Bq·m−3, 162.3 Bq·m−3, and 150.8 Bq·m−3, respectively, across the three periods. Up to 9.5% of the monitored households recorded radon levels exceeding the safety threshold of 200 Bq·m−3. Among the interpolation methods, Kriging provided a more accurate spatial representation of radon concentration variability compared to IDW, allowing for the precise identification of high-risk areas. This study provides a framework for using advanced statistical and geospatial techniques in environmental risk assessment.

The perception of animal emotions is key to enhancing veterinary practice, human–animal interactions, and protecting domesticated species’ welfare. This study presents a unique emotion classification deep learning-based approach for pet animals. The actual and emotional status of dogs and cats have been classified using a modified EfficientNetB5 model. Utilizing a dataset of images classified into four different emotion categories—angry, sad, happy, and neutral—the model incorporates sophisticated feature extraction methods, such as Dense Residual Blocks and Squeeze-and-Excitation (SE) blocks, to improve the focus on important emotional indicators. The basis of the second strategy is EfficientNetB5, which is known for providing an optimal balance in terms of accuracy and processing capabilities. The model exhibited robust generalization abilities for the subtle identification of emotional states, achieving 98.2% accuracy in training and 91.24% during validation on a separate dataset. These encouraging outcomes support the model’s promise for real-time emotion detection applications and demonstrate its adaptability for wider application in ongoing pet monitoring systems. The dataset will be enlarged, model performance will be enhanced for more species, and real-time capabilities will be developed for real-world implementation.

Roads are critical components of infrastructure, and assessing their quality is essential to ensure the safe transport of people and goods, which in turn supports economic prosperity. Various factors, such as subsurface conditions, moisture content, and temperature, influence road performance and can degrade their efficiency as transportation networks. While surface road defects can often be identified through visual inspection, information about subsurface extensions, their impact on structural integrity, and potential risks remain concealed. This study aimed to perform a comparative analysis of dielectric permittivity (ε) using time-lapse Ground Penetrating Radar (GPR) measurements on pre- and post-renovated road sections. This study also sought to evaluate the effectiveness of this approach for road assessment and to employ forward modeling for a deeper understanding of road defects and their associated hazards. Results revealed that the pre-renovated road section exhibited significant fluctuations in dielectric values, ranging from 3.13 to 15.9. In contrast, the post-renovated section showed consistent values within a narrow range of 5 to 6.6. Different crack types were classified, and the mean ε for each visually identified crack type was calculated. Despite the higher frequency of transverse cracks compared to other defects, longitudinal cracks exhibited the highest mean dielectric value (~10.3), while alligator cracks had the lowest (~8.33). Numerical simulations facilitated accurate interpretation of the defects identified in the road section, providing insights into their nature and associated risks. The methodology used for crack classification and numerical simulation can be applied to other road sections globally, offering a standardized approach to road assessment and maintenance planning.

This study explores the feasibility of preparing a one-part alkali-activated binder produced from tungsten-molybdenum (W-Mo) tailings with sodium metasilicate (SM). A series of alkali-activated mortar samples were prepared, and the effects of the water/binder (W/B) ratio and mixture proportion on mechanical properties were investigated. Additionally, the microstructure and composition of the alkali-activated W-Mo tailings were characterized by using a combination of scanning electron microscopy (SEM), X-ray diffraction (XRD), and Fourier transform infrared (FTIR) spectroscopy techniques. Optimal results were achieved with a W/B ratio of 0.35 and a formulation containing 20% by weight of SM. Under these conditions, the cured samples exhibited an unconfined compressive strength of 11.2 MPa and a bulk density of 1726 kg/m3 after 28 days. The findings show the potential to advance tungsten-molybdenum mine waste upcycling and contribute to the production of environmentally sustainable building materials.

End-to-end traceability offers significant opportunities for product lifecycle visibility, sustainability enhancement, and regulatory compliance in product management. However, it faces challenges in data integration and management, supplier collaboration, cost and complexity, and the sharing of information across the supply chain. Productization refers to the representation of a product and connects commercial and technical aspects to the systemic perspective of product management. This includes a focus on the engineering lifecycle with inherent linkages to product data. The product management perspective, specifically in relation to the connection between end-to-end traceability and the productization concept, has not been extensively studied. This study explores the role of both productization and traceability in the context of end-to-end traceability. It combines an extensive literature review and an empirical example of applying productization logic across company borders to support end-to-end traceability. The key findings indicate that productization logic with a product structure focus can support end-to-end traceability in product management by providing consistency and a foundation for tracking both technical and operational data across the engineering lifecycle of a product. By focusing on productization, companies can overcome traceability challenges and unlock the benefits of end-to-end traceability.