This article proposes a deep learning–based autonomous concrete crack detection technique using hybrid images. The hybrid images combining vision and infrared thermography images are able to improve crack detectability while minimizing false alarms. In particular, large-scale concrete-made infrastructures such as bridge and dam can be effectively inspected by spatially scanning the unmanned vehicle–mounted hybrid imaging system including a vision camera, an infrared camera, and a continuous-wave line laser. However, the expert-dependent decision-making for crack identification which has been widely used in industrial fields is often cumbersome, time-consuming, and unreliable. As a target concrete structure gets larger, automated decision-making becomes more desirable from the practical point of view. The proposed technique is able to achieve automated crack identification and visualization by transfer learning of a well-trained deep convolutional neural network, that is, GoogLeNet, while retaining the advantages of the hybrid images. The proposed technique is experimentally validated using a lab-scale concrete specimen with cracks of various sizes. The test results reveal that macro- and microcracks are automatically visualized while minimizing false alarms.

This article proposes a deep learning‐based automated crack evaluation technique for a high‐rise bridge pier using a ring‐type climbing robot. First, a ring‐type climbing robot system composed of multiple vision cameras, climbing robot, and control computer is developed. By spatially moving the climbing robot system along a target bridge pier with close‐up scanning condition, high‐quality raw vision images are continuously obtained. The raw vision images are then processed through feature control‐based image stitching, deep learning‐based semantic segmentation, and Euclidean distance transform–based crack quantification algorithms. Finally, a digital crack map on the region of interest (ROI) of the target bridge pier can be automatically established. The proposed technique is experimentally validated using in situ test data obtained from Jang–Duck bridge in South Korea. The test results reveal that the proposed technique successfully evaluates cracks on the entire ROI of the bridge pier with precision of 90.92% and recall of 97.47%.

Recently, the application of the BIM technique to infrastructure lifecycle management has increased rapidly to improve the efficiency of infrastructure management systems. Research on the lifecycle management of infrastructure, from planning and design to construction and management, has been carried out. Therefore, a systematic review of the literature on recent research is performed to analyze the current state of the BIM technique. State-of-the-art techniques for infrastructure lifecycle management, such as unmanned robots, sensors and processing techniques, artificial intelligence, etc., are also reviewed. An infrastructure BIM platform framework composed of BIM and state-of-the-art techniques is then proposed. The proposed platform is a web-based platform that contains quantity, schedule (4D), and cost (5D) construction management, and the monitoring systems enable collaboration with stakeholders in a Common Data Environment (CDE). The lifecycle management methodology, after infrastructure construction, is then completed and is developed using state-of-the-art techniques using unmanned robots, scan-to-BIM, and deep learning networks, etc. It is confirmed that collaboration with stakeholders in the CDE in construction management is possible using an infrastructure BIM platform. Moreover, lifecycle management of infrastructure is possible by systematic management, such as time history analysis, damage growth prediction, decision of repair and demolition, etc., using a regular inspection database based on an infrastructure BIM platform.

This paper presents a deep learning-based concrete crack detection technique using hybrid images. The hybrid images combining vision and infrared (IR) thermography images are able to improve crack detectability while minimizing false alarms. Large scale concrete-made infrastructures such as bridge, dam, and etc. can be effectively inspected by spatially scanning the hybrid imaging system including vision camera, IR camera and continuous-wave line laser. However, the decision-making for the crack identification often requires experts’ intervention. As a target concrete structure gets larger, automated decision-making becomes more necessary in the practical point of view. The proposed technique is able to achieve automated crack identification by modifying a well-trained convolutional neural network using a set of crack images as a training image set, while retaining the advantages of hybrid images. The proposed technique is experimentally validated using a lab-scale concrete specimen developed with various-size cracks. The test results reveal that macro- and micro-cracks are automatically detected with minimizing false-alarms.

This paper proposes a self-sensing nonlinear ultrasonic technique for fatigue crack detection under temperature variations. Fatigue cracks are identified from linear (α) and nonlinear (β) ultrasonic parameters recorded by a self-sensing piezoelectric transducer (PZT). The self-sensing PZT scheme minimizes the data acquisition system’s inherent nonlinearity, which often prevents the identification of fatigue cracks. Also, temperature-dependent false alarms are prevented based on the different behaviors of α and β. The proposed technique was numerically pre-validated with finite element method simulations to confirm the trends of α and β with changing temperature, and then was experimentally validated using an aluminum plate with an artificially induced fatigue crack. These validation tests reveal that fatigue cracks can be detected successfully in realistic conditions of unpredictable temperature and that positive false alarms of 0.12% occur.

This study presents the development of a paint condition evaluation (PCE) system that integrates a wall-crawling drone with a laser thermography module for the remote and rapid inspection of steel bridge paint conditions. A wall-crawling drone was employed to access hard-to-reach areas for inspection. The drone crawled along the inspection surface when it arrived at the target inspection area. Simultaneously, the laser thermography module heats the target surface using a laser beam, and the resulting heat responses are recorded using an infrared (IR) camera. The recorded heat responses were analyzed using the PCE algorithm developed in this study, enabling visualization and quantification of the paint thickness distribution and hidden delamination. The key contributions of this study include (1) the integration of laser thermography with a drone for rapid and remote PCE, (2) inspection of inaccessible areas within a bridge structure using a drone with motorized wheels, and (3) automated quantification and visualization of invisible paint thickness distribution and hidden delamination. The performance of the PCE system was validated through lab-scale and field-bridge tests, demonstrating the successful visualization of the paint thickness and delamination area. The quantification error was less than 10 μm for paint thickness and 1 cm ² for delamination area.