This study presents an integrated framework for bridge inspection that combines multiple non-destructive testing (NDT) technologies with artificial intelligence (AI) and immersive visualization. The proposed system integrates, and fuses unmanned aerial vehicles (UAV)-based LiDAR point clouds, photogrammetry, infrared thermography (IRT), and phased-array ultrasonic tomography (UT) to generate a comprehensive, bridge-scale 3D inspection model with specific application to bridge decks. A fine-tuned Grounding DINO object detection model, trained on 10,500 infrared images, is used to automatically identify suspicious thermal patterns. The AI achieved 90% precision, 90% recall, an F1 score of 0.90, and a mean average precision (mAP@0.5) of 0.80 on held-out test data. These detections are exported as geo-referenced waypoints to guide targeted UT scans, which confirm and characterize subsurface defects such as delamination and voids. All sensing outputs are aligned within a unified coordinate system and visualized inside a virtual reality (VR) environment. Users can interact with 3D geometry, thermal overlays, and depth-resolved UT slices, and annotate defects in context. By replacing manual IRT interpretation and full-grid UT scanning with AI-guided anomaly detection and selective validation, the proposed workflow has the potential to reduce inspection time, lowers labor costs, and minimizes subjectivity in data interpretation. This system also provides a centralized, interactive 3D record that supports efficient decision-making and long-term maintenance planning.

Digital twins (DTs) are increasingly seen as a promising tool for bridge management, yet bridge owners hold diverging views on bridge digital twin (BDT) requirements, benefits, and feasibility. The International Association for Bridge Maintenance and Safety (IABMAS) Working Group on Digital Twins conducted an international survey of bridge owners and reviewed current developments in DT and case studies. The survey covered owners’ management needs, DT definitions, input/output expectations, and perceived potential. Owners see promise in DTs but want clearer guidance on minimum requirements, practical benefits, and scalable implementation strategies. Case studies illustrate their strengths in lifecycle decision-making and weaknesses tied to data fragmentation. Coherent data standards, platform interoperability, and integration with existing asset management workflows are essential. This work outlines a roadmap toward a harmonized, owner-centered BDT framework grounded in coherent standards, scalable implementation paths, and cross-sectoral collaboration.

Ensuring the safety and longevity of bridge decks demands effective inspection techniques, yet traditional methods often rely on manual analysis and time-intensive procedures. This paper introduces a streamlined, AI-driven approach that combines vehicle-mounted infrared (IR) imaging, a Transformer-based detection model, and ultrasonic tomography (UT) to identify and assess deck anomalies. First, processed IR scans are used to generate labels for the raw IR data, enabling the AI model to detect suspicious regions directly from minimally processed imagery. Potential defect areas, including delaminations and cracks, are then verified through UT, providing three-dimensional insights into their depth and extent. By leveraging an AI model that demonstrates a 70% accuracy when compared to processed IR ground truth, this integrated workflow reduces the need for extensive manual preprocessing, accelerates inspection, and delivers precise global localization of defects. Overall, the proposed methodology offers a scalable solution that improves both inspection reliability and maintenance planning for critical bridge infrastructure.

Accurate geometric and structural data are essential for the development of digital twins for existing infrastructure systems, particularly in the context of building resilience. Many older or poorly documented structures either lack as-built plans or exhibit deviations from the original design due to construction inconsistencies, undocumented modifications, or repairs. As a result, capturing current structural conditions becomes a necessary first step for generating reliable analytical models, conducting assessments, and enabling long-term infrastructure monitoring. In this study, we evaluate the effectiveness of LiDAR-derived point clouds in generating finite element models (FEM) of a infrastructure, using a stadium as a case study. We compare models built from structural plans and point cloud data against on-site measurements in terms of geometry, dimensions, joint displacements, and support reactions. Our results show that point cloud-based models consistently outperform plan-based models in replicating actual geometric and structural behavior. These findings demonstrate that point cloud scanning can serve as a reliable, efficient, and scalable method for generating digital structural models. This capability is especially valuable for building resilient infrastructure systems, where timely and accurate digital representations are needed for performance tracking, diagnostics, and model updating over time. The study highlights the practical application of point cloud data for digital twin development in large-scale, load-bearing public structures.

This paper presents an approach to enhance the implementation of the Bridge Information Modeling (BrIM) methodology during the operational stage by automating the integration of inspection data from Bridge Management Systems (BMS) into BrIM models. While the data from BMS is available and retrievable from spreadsheets, the 3D bridge model is represented according to the Industry Foundation Classes (IFC) data model. Then, this paper introduces an algorithm that ensures seamless interoperability between the spreadsheets and the IFC data model. The semantically enriched BrIM models are achieved through a set of rules and procedures that are established to simplify the matching of the modeled objects with the components that characterize the bridge in the BMS. The openness of the IFC data model allows the identification of appropriate entities to store the corresponding information that comes from the BMS. This automated process removes the need for manual attachment of inspection data into IFC files, which is prone to errors, reduces the complexity of moving towards BrIM-based bridge management practices, and increases the efficiency of creating BrIM models for existing bridges. Finally, the proposed approach is a scalable and transferable solution that transportation agencies worldwide can adopt to manage bridge assets.

Ultrasonic tomography is a powerful nondestructive technique for evaluating internal defects in concrete structures. This study presents a deep learning–enhanced approach utilizing a nanoscale object detection model to automate the localization and quantification of internal defects and embedded structural components, including reinforcement bars and ducts. Controlled concrete samples containing artificial defects of varying shapes and depths, along with embedded rebars and ducts, were designed. Ultrasonic signals were collected using a MIRA A1040 tomograph and reconstructed into 3D volumes via Synthetic Aperture Focusing Technique (SAFT). These volumes were converted into 2D slices and segmented using Chan-Vese segmentation and morphological post-processing. A partial histogram matching procedure unified color scales across segmented slices, minimizing color-related biases before model training. Segmentation-assisted labeling provided robust ground truth annotations, resulting in 7220 labeled images. The trained AI model accurately detected delaminations, rebars, and ducts (both grouted and ungrouted), achieving a mean Average Precision (mAP@0.5) of 0.73 and an Average Intersection-over-Union (IoU) of 0.80. Testing on real-world bridge data demonstrated the model's generalization to unseen conditions. Key innovations include automated segmentation-based labeling, robust color standardization via histogram matching, and a lightweight deep learning model optimized for real-time deployment on resource-constrained devices. This integrated approach has the potential to reduce manual interpretation and subjective variability, providing an effective, scalable NDT/E solution for rapid assessment and monitoring of concrete infrastructure through advanced ultrasonic imaging combined with standardized, machine learning-based defect detection.

In this study, a novel approach using a modified time series analysis methodology is used to detect, locate, and quantify structural changes by using ambient vibration data. Random Decrement (RD) is used to obtain pseudo free response data from the ambient vibration time histories. ARX models (Auto-Regressive models with eXogenous input) are created for different sensor clusters by using the pseudo free response of the structure. The output of each sensor in a cluster is used as an input to the ARX model to predict the output of the reference channel of that sensor cluster. After the ARX models for the healthy structure for each sensor cluster are created, the same models are used for predicting the data from the damaged structure. The difference between the fit ratios is used as damage indicating feature. The methodology is applied to experimental data obtained from steel grid structure tests and it is shown that the approach is successfully used for identification, localization, and quantification of different damage cases for both impact and ambient tests.

Traditional bridge deck inspections often involve manual labour and data recording, which can be time-consuming and error-prone. Infrared (IR) has strong potential to improve inspections. However, manual IR data processing can also be time-consuming. This paper presents a methodology that integrates IR imaging, an Artificial Intelligence (AI) model and Ultrasound Tomography (UT) to streamline and enhance inspections. The approach begins with vehicle-mounted IR imaging for rapid, large-scale scanning of bridge decks to identify potential concerns. Unlike conventional methods and earlier AI-integrated studies relying on pre-processed IR data, this approach uses processed IR data to label unprocessed images, which are then used to train a Grounding DINO AI model. The trained model autonomously detects and localises suspicious regions directly from raw IR images, eliminating labour-intensive processing and enabling real-time defect detection. UT is subsequently employed to provide detailed analysis of flagged areas, offering insights into damage types such as delamination, spalling and defect depth. The AI model's precision, around 90%, is evaluated against ground truth from processed IR data. Integrating these technologies, the proposed method offers great potential to accelerate inspections, improve defect localisation with global coordinates and support efficient maintenance planning.

This study explores the potential applications of blockchain technology to enhance data authenticity and security in real-time monitoring systems. Considering the building or the structure (e.g., bridge, building frame), both systems can be monitored via IoT. The collected data enable the construction of an information model (digital shadow). Such conceptual models derived from the data can then be certified via blockchain, on which smart contracts can be set up. From this approach, two different strategic paths can be highlighted in the area of efficiency and security. The first is related to the structural domain and concerns the identification of emergency situations through automatic algorithmic processing of the raw data collected by sensors: if IoT identifies the exceeding of a safety threshold, for example, thus associated with abnormal behavior or damage, this condition is mapped and certified in real time. The smart contract set on the blockchain can at this point automatically arrange for intervention on the emergency condition (operators warning, interventions, recovery actions, etc.). On the other hand, the certification of the raw data, in parallel with the certification of the processed data, via blockchain allows for the reconstruction of the emergency condition from the perspective of scenario but also in terms of legal implications. The same blockchain and smart contract associative process can be translated from the emergency scenario to the service scenario for the management of routine maintenance contracts.

Incorporating non-destructive evaluation (NDE) techniques into bridge assessment decision-making offers significant advantages to bridge owners. One promising method, infrared thermography (IRT), has been investigated for detecting concrete anomalies in civil engineering. When combined with high-definition image scans, IRT shows potential for rapid and efficient bridge scanning applications. This paper presents a comparative analysis of two rapid bridge scans conducted using a combination of IRT and high-definition imaging (IRT+Image). These scans were performed in 2021 and 2022 on a deteriorating bridge (built in 1964). The results of the scans are then compared with the findings from the official biennial visual inspection of the bridge. Additionally, a decision-making model based on a perceptron neural network is developed and demonstrated, which leverages the scan results. This model can be expanded to incorporate other NDE techniques or visual inspection data within an element-level inspection framework. The comparative evaluation showed that the majority of the concrete defects identified in the inspection report were also identified in the IRT+Image scans. Furthermore, the IRT+Image scan identified additional areas of concrete defects that were not identified in the inspection report. However, these additional locations were not verified to represent defects by means other than the scan. For example, IRT+Image scan identified 42.5% more piles with delamination than the inspection did. For the deck underside, 99% of the defect locations identified in the visual inspection were also detected by the IRT+Image scan. For the superstructure beams, the correspondence was 79%, and for the scanned bents, all substructure piles with defects in the biennial inspection were similarly identified by the IRT+Image scan. The perceptron neural network developed was applied to two case studies, assisting in decision-making for further evaluation. This practical NDE application using IRT+Image and the decision-making model can be extended to other transportation assets and integrate various NDE technologies.

This study investigates the capture of digital data and the development of models for structures with incomplete documentation and plans. LiDAR technology is utilized to obtain the point clouds of a pedestrian bridge structure. Two different point clouds with varying densities, (i) fine (11 collection locations) and (ii) coarse (4 collection locations), collected via terrestrial LiDAR, are analyzed to generate geometry and structural sections. This geometry is compared to the structural plans, which are then converted into numerical models (finite element—FE model) based on the point cloud data. Point cloud-based FE models (based on fine and coarse data) are compared with the structural plan-based FE model. It is observed that the static and dynamic responses are comparable within an acceptable range of a maximum difference of 5.5% for static deformation and an 8.23% frequency difference, with an average difference of less than 5%. Additionally, the dynamic properties of the fine and coarse point cloud FE models are compared with the operational modal analysis data obtained from the bridge. The fine and course point-cloud-based FE models, without model calibration, achieve an average accuracy of 8.76% and 9.94% for natural frequencies and a 0.89 modal assurance criterion value. The research found that the digital data generation yields promising results in this case for a bridge if documentation or plans are unavailable. With recent technologies and approaches such as digital twins, the connection between physical and virtual entities needs to be established by fusing digital models, sensorial information, and other data forms for better infrastructure management. Models such as those investigated and discussed in this paper can assist engineers with structural preservation in conjunction with monitoring data and utilization for digital twins.

When the load rating of a bridge is less than 1.0 for AASHTO HL-93 live load and state-specific legal trucks, the bridge is posted. Posting a bridge causes an inconvenience to the public and may result in trucks taking longer routes. Thus, this paper investigated the effects of field-derived distribution factor (DF) and impact factor (IM) from static and dynamic load tests using computer vision and deflection measurements, instead of AASHTO factors, on the bridge load rating and component as well as system reliability for posting avoidance. The reliability approach used Monte Carlo simulations to account for uncertainties in calculating the flexural strength limit state (Strength I). Both Flexural strength I and serviceability (Service III) limit states were investigated. The bridge's superstructure was studied in as-built, repaired, and damaged condition scenarios. For that, AASHTO HL-93, Florida legal trucks, and emergency vehicles were considered. In this regard, load rating distributions and component and system reliability indices were computed using both AASHTO and field-derived DF and IM. The increase in load rating and reliability was investigated. By using field-derived factors, a load rating increase of up to 17% was achieved. The component reliability also increased significantly. For example, the damaged case's component reliability increased by 1.0 when replacing DF and IM, which indicates a major reduction in the probability of failure. The increase in system reliability was most significant because using the field-derived DF increased the capacity contribution from other girders. A 2.12 increase in system reliability was achieved in the damaged condition when only substituting DF, providing even higher system reliability as a result of a reduction in the system probability of failure.

Dynamic ultrasonic waves can be utilized to assess concrete for defects. This study introduces an optimized approach using computer vision algorithms on ultrasound images to assist in localizing and quantifying substantial defects and components. To detect concrete details, the ultrasonic signals are assembled to create 3D images for the concrete material and then converted into 2D so-called B-scans. A computer vision algorithm was trained on a dataset of labeled images that show different levels of delamination as 2D ultrasound images. The algorithm can then analyze new ultrasound images and classify them according to the degree of delamination present. The computer vision techniques that can be used for this purpose include convolutional neural networks (CNN) by applying YOLO V5, commonly used for image classification tasks. In addition to detecting delamination, computer vision techniques can identify other abnormalities in concrete structures, such as cracks, voids, and inclusions. These techniques can help engineers and maintenance professionals to identify and address problems in concrete structures before they lead to more severe issues. The paper will present these methodologies and applications with a case study.

In this study, an existing six-storey reinforced concrete building with an asymmetric structural plan and soft storey irregularity was used as a test specimen and subjected to three-step progressive structural damages to detect the variations in its dynamic properties. Mode shapes and dominant frequencies of the undamaged building were determined by the ambient vibration survey (AVS) and it was seen that its first three modes were torsion coupled. Besides, soft storey irregularity was evident due to the lack of masonry infill walls on its ground floor. Later on, three-step progressive damages were applied to the building. The first step targeted three columns and one beam of the building, located on a corner region of its ground floor to peel off their clear covers. The second step razed two adjacent corner columns which were already moderately damaged in the first step, while the third step knocked the third moderately damaged column down. After each damage step, AVS was repeated with the same details as applied for the undamaged building. The obtained dynamic properties for the four phases of the building were evaluated with the sustained damage. Numerical analyses with the finite element model of the building representing its four different phases were also performed and the unique responses due to damage effects on the structure were investigated numerically. As a result of induced damage, the quantified amount of frequency change in modes and the new mode observed after particularly column loss scenarios can be utilized for efficient structural health-monitoring strategies of plan-asymmetric buildings and post-earthquake assessment of partially damaged buildings where timely objective assessment is important.

A review of the application of remote sensing technologies in the SHM and management of existing bridges is presented, showing their capabilities and advantages, as well as the main drawbacks when specifically applied to bridge assets. The main sensing technologies used as corresponding platforms are discussed. This is complemented by the presentation of five case studies emphasizing the wide field of application in several bridge typologies and the justification for the selection of the optimal techniques depending on the objectives of the monitoring and assessment of a particular bridge. The review shows the potentiality of remote sensing technologies in the decision-making process regarding optimal interventions in bridge management. The data gathered by them are the mandatory precursors for determining the relevant performance indicators needed for the quality control of these important infrastructure assets.

This paper introduces a Virtual Meeting Environment (VME) leveraging Virtual Reality and investigates its impact on decision-making during infrastructure inspections. The investigation involves practical experiments with structural engineers using VME, followed by a questionnaire and simulations in the Inspection Simulation Environment (ISE) to obtain decision-making results. The questionnaire results show that 91% of the engineers believe using VME as an additional tool to on-site and video conference meetings, which is the current practice, improves meeting efficiency and assists in decision-making in infrastructure inspections. Simulation results demonstrate that using VME alongside current practices improves decision-making, as evidenced by a reduction in the mean Decision Variance (DV) from 5.11% to 3.56%. This improvement is even more pronounced when compared to solely on-site meetings, where the mean DV decreases from 8.05% to 3.56%. The paper shows that the more reliable the data collection and analysis tools are, the further the VME improves decision-making.

The stadium structures have unique structural features increasing the significance of structural monitoring systems specifically designed for them. Aside from vibrations serviceability concerns and human-induced excitations, the development and propagation of structural damage under all possible atmospheric and seismic conditions need to be closely monitored for structural resiliency and integrity of the stadia. As such, Structural Health Monitoring (SHM) methods combined with effective data evaluation methodologies need to be deployed to monitor the structural performance of stadiums. Even though stadia monitoring has been performed at multiple locations in the world, a web based and real-time SHM network of stadia is not known to authors. As a preliminary study for the network implementation of stadia monitoring with acceleration measurements, the presented work focuses on the fundamental steps to accomplish this goal, with a collaborative research effort between Qatar University, the University of Central Florida, and University of Alberta. The authors performed analytical investigations and experimental testing on stadium-type structures built in laboratory environments for the development of the SHM framework. Specialized signal processing algorithms, sensing suites and approaches considering multi-scale monitoring were used on collected acceleration measurements. The novelty of the work presented in this manuscript are the following items which exist simultaneously in the developed SHM framework. The developed framework is a web-based monitoring application where structural damage is detected in real-time. The proposed methodology operates directly on raw acceleration signals and runs at a network level. With that, the damage detection, damage localization, and damage quantification tasks are performed simultaneously, while the feature extraction and classification stages are combined in one learning body.

This paper presents a unique approach to data fusion in the condition assessment of civil infrastructure systems. The approach fuses dynamic monitoring sensor data with the visual data collected from a bridge structure. Such a data fusion method, named ModeShapeFuser, enables the visualization of mode shapes of engineering structures in their point cloud, offering a higher spatial resolution model of the dynamic behaviors of existing structures. A bridge case study shows that utilizing this immersive data visualization technique can potentially provide a more intuitive and holistic understanding of the dynamic behavior characteristics of engineering structures. The outputs of ModeShapeFuser, the high spatial resolution models of the bridge, are also presented in a Virtual Reality (VR) environment through a head-mounted display and a VRTable, providing a virtual tour of the bridge in its real environment. The study presented herein is critical for an informed decision-making process in assessing infrastructure systems.