A Data-Driven Based Response Reconstruction Method of Plate Structure with Conditional Generative Adversarial Network.

Structural-response reconstruction is of great importance to enrich monitoring data for better understanding of the structural operation status. In this paper, a data-driven based structural-response reconstruction approach by generating response data via a convolutional process is proposed. A conditional generative adversarial network (cGAN) is employed to establish the spatial relationship between the global and local response in the form of a response nephogram. In this way, the reconstruction process will be independent of the physical modeling of the engineering problem. The validation via experiment of a steel frame in the lab and an in situ bridge test reveals that the reconstructed responses are of high accuracy. Theoretical analysis shows that as the sensor quantity increases, reconstruction accuracy rises and remains when the optimal sensor arrangement is reached.

Investigation of the Internal Structure of Hardened 3D-Printed Concrete by X-CT Scanning and Its Influence on the Mechanical Performance.

As we know, 3DPC is printed layer by layer compared with mold-casting conventional concrete. Pore structure and layer-to-layer interface are two main aspects of the internal structure for 3DPC, which decide 3DPC's mechanical performance. The layer-to-layer interface caused by printing is specific to 3DPC. The emphasis of this study lies in the layer-to-layer interfaces of 3DPC. The first aim of this study is to quantify the characteristics of the layer-to-layer interface and therefore characterize different aspects of the interfaces. The second aim of this study is to explore how the internal structure of printed concrete influences the mechanical performance of 3DPC. This research set out to design a series of experimental comparisons between 3DPC and casted concrete with the same compositions. Mechanical tests, i.e., compressive stress, ultrasonic Pulse Velocity test, flexural tension, and tension splitting, as well as the Ultrasonic Pulse Velocity test, were performed to check the mechanical performance of 3DPC. Contrary to what has often been expected, the mechanical test results showed the printed concrete has a quality not worse than casted concrete with the same recipe. Meanwhile, the X-ray computed tomography (X-CT) is used to characterize the internal structure, pore shapes, and interfaces of 3DPC. First, the investigation revealed that the lower total porosity and fewer big voids could be the fundamental causes meaning 3DPC has a better mechanical performance than casted concrete. Second, the statistics based on aspect ratio show that the distribution curves follow similar trends, regardless of the printed or casted concrete. Third, this study quantified the depth of the different interfaces for 3DPC. The results suggest that the porosity in an interface varies in a range. The author's pioneer work has contributed to our present understanding of the interfaces of 3DPC.

A Binocular Vision-Based Crack Detection and Measurement Method Incorporating Semantic Segmentation.

The morphological characteristics of a crack serve as crucial indicators for rating the condition of the concrete bridge components. Previous studies have predominantly employed deep learning techniques for pixel-level crack detection, while occasionally incorporating monocular devices to quantify the crack dimensions. However, the practical implementation of such methods with the assistance of robots or unmanned aerial vehicles (UAVs) is severely hindered due to their restrictions in frontal image acquisition at known distances. To explore a non-contact inspection approach with enhanced flexibility, efficiency and accuracy, a binocular stereo vision-based method incorporating full convolutional network (FCN) is proposed for detecting and measuring cracks. Firstly, our FCN leverages the benefits of the encoder-decoder architecture to enable precise crack segmentation while simultaneously emphasizing edge details at a rate of approximately four pictures per second in a database that is dominated by complex background cracks. The training results demonstrate a precision of 83.85%, a recall of 85.74% and an F1 score of 84.14%. Secondly, the utilization of binocular stereo vision improves the shooting flexibility and streamlines the image acquisition process. Furthermore, the introduction of a central projection scheme achieves reliable three-dimensional (3D) reconstruction of the crack morphology, effectively avoiding mismatches between the two views and providing more comprehensive dimensional depiction for cracks. An experimental test is also conducted on cracked concrete specimens, where the relative measurement error in crack width ranges from -3.9% to 36.0%, indicating the practical feasibility of our proposed method.

Collision-free trajectory planning for robotic assembly of lightweight structures.

This research presents a trajectory planning approach for robotic assembly of lightweight structures for COVID-19 healthcare facilities. The prefabricated building components of COVID-19 healthcare facilities have nonnegligible volume, where the crux of the scientific question lies in how to incorporate geometry-based collision checks in trajectory planning. This research developed an algorithm that refines the RRT* (Rapidly-exploring Random Tree-Star) algorithm to enable the detour of a planned trajectory based on the geometry of prefabricated components to prevent collisions. Testing of the approach reveals that it has satisfactory collision-avoiding and trajectory-smoothing performance, and is time- and labour-saving compared with the traditional human method. The satisfactory results highlight the practical implication of this research, where robots can replace human labour and contribute to the mitigation of COVID-19 spread on construction sites. The subsequent research will investigate the use of a collaborative robot to screw bolt connections after the components are assembled at locations.

BIM-based task and motion planning prototype for robotic assembly of COVID-19 hospitalisation light weight structures.

Fast transmission of COVID-19 led to mass cancelling of events to contain the virus outbreak. Amid lockdown restrictions, a vast number of construction projects came to a halt. Robotic platforms can perform construction projects in an unmanned manner, thus ensuring the essential construction tasks are not suspended during the pandemic. This research developed a BIM-based prototype, including a task planning algorithm and a motion planning algorithm, to assist in the robotic assembly of COVID-19 hospitalisation light weight structures with prefabricated components. The task planning algorithm can determine the assembly sequence and coordinates for various types of prefabricated components. The motion planning algorithm can generate robots' kinematic parameters for performing the assembly of the prefabricated components. Testing of the prototype finds that it has satisfactory performance in terms of 1) the reasonableness of assembly sequence determined, 2) reachability for the assembly coordinates of prefabricated components, and 3) capability to avoid obstacles.