This PDF file contains the front matter associated with SPIE Proceedings Volume 12952, including the Title Page, Copyright information, Table of Contents, and Conference Committee information.

This paper recommends guidelines for implementing supplier nondestructive evaluation (NDE) qualification and certification program intended for inspecting fracture critical parts. Qualification of fracture critical parts may require qualified NDE procedures that assure reliably detectable flaw size. It is intended for NDE procedures that detect flaws or discontinuities. Some guidelines for implementing NDE qualification and certification for NDE procedures that assure reliably detectable flaw size are available in current literature for research publications or as industry standards. This paper proposes additional considerations especially for qualifying NDE procedures, establishing a written practice for NDE procedure, including personnel qualification and certification. The supplier NDE qualification and certification implementation practice is referred to as the supplier NDE qualification and certification program. The paper touches upon many elements of NDE as a systems engineering discipline. If supplier qualification and certification program is less rigorous or has gaps, it may lead to lower reliability in flaw detection. These guidelines are intended to help in acquiring minimum required reliability in supplier NDE qualification and certification. The guidelines are applicable for all NDE procedures including those assuring reliably detectable flaw sizes. Before recommending guidelines, the paper defines core topics such as NDE reliability, probability of detection, NDE transfer function, resolution, contrast sensitivity and points to relevant references.

In the production and packaging of silicone sealants, entrapped air, impurities, and foreign particles can introduce defects affecting performance. This study utilizes three ultrasonic non-destructive testing techniques: contact-based, angle-beam, and through-transmission testing, to identify defects and generate 3D images. The scanning process takes vertical A-scan measurements across the sample and rotates it at predetermined angles for comprehensive coverage. The contact-based technique uses the time-of-flight principle to determine defect locations, but struggles with defects aligned perpendicular and experiences signal reduction. The angle-beam method identifies defects in areas previously out of reach, but the slow sound movement in sealants can hinder capturing specific signals. While through-transmission offers enhanced signal clarity and an improved signal-to-noise ratio, pinpointing the defect’s exact depth is challenging. By combining these methods, the study reconstructs a more accurate three-dimensional image which visualizes the defective region.

In the inspection of cracks by structural health monitoring system, it is common to acquire 2D images. However, when the target is a thick crack, it is difficult to identify the width of the crack, especially if the positional relationship between the lighting and the camera is not sufficiently taken into account. On the other hand, since infrastructure inspection vehicles are being put into practical use, it is necessary to identify crack widths even with 2D images. Therefore, this study proposes a method to calculate the crack width by image processing using chalk-marks drawn during visual inspection of wide cracks in 2D images acquired by a structural health monitoring system while traveling at high speed. Experiments showed that the proposed method can detect cracks within a margin of error of 10 to 15%, while conventional crack detection software cannot calculate the width.

In contemporary times, there has been a reduction in the length of product lifecycles, accompanied by a growing consumer preference for more complex offers. The reality poses numerous obstacles, and the existing methodologies are not viable in the long term. In recent years, there has been a notable surge in scholarly attention towards the domain of prognostics. Most research efforts have mostly focused on the prediction of the remaining useful life (RUL) of individual components. The dissemination of failure mechanisms can also involve several components, and a range of prognostic methods are utilised to detect and track them at different levels of the system. Once specific thresholds of deterioration have been reached, it becomes crucial to implement specific maintenance measures to mitigate the risk of a potential system failure. Therefore, it is crucial to understand how the mechanisms have and will be distributed while striving to perform certain maintenance methods to extend the RUL. The authors in this research introduce a prognostic methodology for predictive maintenance that relies on the physics of failure (PoF) approach and statistical method. The objective of this methodology is to employ knowledge of aircraft fuel system failure mechanisms and use datasets derived from laboratory experiments and simulation model conducted on aircraft fuel systems. The proposed methodology integrates sensor data and algorithms to assess the degradation of a system from its expected normal operational condition as well as to predict its RUL. The proposed methodology has a high degree of robustness and consistently produces trustworthy outcomes.

Structure from Motion (SfM) is a photogrammetry technique with diverse applications, such as surveying, mapping, and inspection. It facilitates remote assessment of large systems by converting visible spectrum images into threedimensional (3D) point clouds. Recent advancements have extended SfM to employ infrared (IR) images, enabling the detection of issues such as water infiltration and sub-surface defects that can cause energy loss in a building. Combining IR-based SfM with unmanned aerial vehicle (UAV) technologies yields high-definition 3D point clouds that can be used in a virtual reality (VR) environment. This study showcases the application of the SfM-IR-UAV method to create 3D virtual models of selected buildings in the University of Massachusetts Lowell’s campus to assess energy loss. The 3D virtual models are made accessible via a VR platform to develop a remote inspection and maintenance tool. The VR platform also holds the capabilities to mark abnormalities in the structure, which can later be used for informing renovation or repair. The proposed approach simplifies remote assessment, reducing costs and operational risks. While this research focuses on energy audits, its outcomes extend to diverse domains. Further development holds the potential to expedite nondestructive evaluation and enhance structural health monitoring in civil and mechanical engineering, utilizing the 3D point cloud thermal model within a VR environment.

Building energy consumption grows rapidly with modern urbanization while the buildings’ sensor data also increases explosively. Improving energy utilization of community buildings is critical for sustainable development and global climate challenge. However, the data isolation across buildings’ privacy management prevents largescale machine learning model training, which may reduce the prediction accuracy due to lack of data. Federated building energy learning supports distributed learning through model sharing so that data privacy is mitigated. In federated learning, model-sharing brings a new concern about network resource limitation. Deep learning model transfers across multiple buildings would cause network ingestion and incur high latency of federated training. To improve the efficiency of federated training with fewer resources, a new federated learning algorithm is proposed with a new deep learning model design. The deep learning model memory usage is reduced by 80% while energy load forecasting accuracy is still comparable to the state-of-the-art methods.

Recent research has mostly investigated the theoretical performance of various structural acoustic monitoring systems, but there is little data for these systems’ real-world performance. Intermittent signal strength, anomalous data loss, and other implementation issues can occur and must be addressed before wide-scale adoption of this technology is possible. Key differences that exist between in-lab testing and real-world implementation that this study addresses are: constant rotational motion, complex blade internal structure, costly cellular data, and remote location. In this paper, we present the design and testing of an acoustic sensor based structural health monitoring system. A field test was conducted to evaluate the performance of antenna candidates. We designed IoT acoustic sensors mounted within wind turbine blades, including a local controller node that aggregates data and pushes them to the cloud via cellular networks, and a cloud dashboard developed on Amazon Web Services. During the six-month (April through October 2022) deployment at National Renewable Energy Laboratory (NREL), acoustic samples as well as other metrics such as temperature, acceleration, wireless signal strength, data transfer latency were collected continuously which enable analysis of the health of the blades and validate the functionality of the monitoring system.

Pipeline systems are critical infrastructure for modern economies, which serve as the essential means for transporting oil, gas, water, and other fluids. These pipelines are mostly buried underground, making their integrity highly crucial. Because they are buried, these pipelines are subject to stress and are prone to material degradation due to corrosion. Corrosion not only reduces the wall thickness of the pipes but also poses severe safety risks and can lead to catastrophic failures and substantial financial losses. Hence, there is an urgent need to develop accurate predictive models for evaluating pipe wall thickness. This paper aims to address this need by exploring machine learning-based algorithms to monitor the corrosion rates so that preventive measures can be taken to ensure pipeline integrity. Thus, four state-of-the-art machine-learning algorithms, namely, Convolutional Neural Network (CNN), Gated Recurrent Unit (GRU), Bidirectional Gated Recurrent Unit (Bi-GRU), and Long Short-Term Memory (LSTM) are employed to predict accurate wall thickness of pipelines. The empirical results show that the LSTM algorithm outperforms its counterparts, achieving a low root mean squared error (RMSE) of 0.0721 mm. Therefore, incorporating LSTM-based models into pipeline integrity programs can be a significant step forward to safeguard these critical infrastructures.

Magnetic flux leakage (MFL) is a widely used nondestructive testing technique in pipeline inspection to detect and quantify defects. In pipeline integrity management, the reconstruction of defects from MFL signals plays a critical role in failure pressure prediction and maintenance decision-making. In current research practices, this reconstruction primarily involves the determination of defect dimensions, including length, width, and depth, collectively forming a rectangular box. However, this box-based representation potentially leads to conservative assessments of pipeline integrity. To fine-scale the reconstruction results and provide detailed defect information for the integrity assessment, a 3-D reconstruction model for pipeline corrosion defects from MFL signals is proposed. In detail, the deep neural network is established to capture the nonlinear relationship between the MFL signals and 3-D defect profiles. In contrast to the limited insights offered by the box profile, the reconstructed 3-D profile in this paper enables more detailed metal loss geometry. The experiments using field pipeline in-line inspection data demonstrate promising results on both morphology and depth prediction.

Energy load forecasting across multiple buildings is beneficial for energy saving. Currently, most methodologies are training a single global model for all buildings as the deep learning model relies on large-scale data. However, the energy data distribution may vary a lot across different buildings and enforcing a global model may cause unnecessary computing resource overutilization. Meanwhile, building energy management encounters repeated manual efforts for machine learning model training over the new sensor data. To improve the computing resource utilization of load forecasting model training and automation of building energy management, a new automatic learning framework is proposed to support automatic building energy data analytics. The machine learning model is customized for each building based on an automatic algorithm with efficient model evaluations. The new framework brings comparable performance to federated energy data learning while fewer computing resource is consumed.