The increasing interest in Digital Twins (DT) for road infrastructure stems from their ability to seamlessly integrate data from various sources and facilitate interactions between physical entities and their virtual counterparts. The cornerstone for achieving a road digital twin is a high-quality road model. Despite the widespread application of Building Information Modeling (BIM) technology across various stages of road engineering, the modeling of existing roads within the BIM framework has faced challenges, including a lack of accurate design data and extensive manual work. This paper presents an innovative and cost-effective approach to model existing roads by combining Uncrewed Airborne Vehicles (UAVs) -LiDAR System (ULS) and BIM. It begins with the development of a low-cost UAV-LiDAR platform designed for efficient data collection in road environments, capturing essential geometric information. Subsequently, point cloud mapping techniques, including an intensity-based planar downsampling strategy and a hierarchical refinement method, are employed to create accurate point cloud maps using the cost-effective ULS. Further, an automated modeling pipeline is proposed with point cloud processing and Revit Dynamo. It comprises three main steps: cross-section model design, point cloud parameter extraction, and the design of an automated workflow in Dynamo. The proposed method is validated with both qualitative and quantitative experiments, and the results demonstrate that the generated model accurately represents the structural, alignment, and lateral gradient details of the experimental road. In summary, this study offers a cost-efficient yet highly effective means to construct realistic models of existing roads.

The accurate representation of pavement in three-dimensional (3D) space is pivotal for road infrastructure management. However, the sparse measurement from affordable equipment, such as 2D LiDAR, has limited the fully utilization of 3D pavement data. This paper aims to address this challenge by exploring the point cloud upsampling for pavement area. Specifically, Pavement-PU, a novel data-driven model is designed to enhance the quality and efficiency of dense point cloud generation task. Utilizing a transformer-based feature extraction module coupled with an integrated point upsampling strategy, Pavement-PU significantly improves the density, uniformity, and accuracy of point clouds derived from sparse and irregular initial scans. Through rigorous testing on the public dataset and a specially curated Pavement3D dataset, the model demonstrates substantial improvements over existing methods in terms of both quantitative metrics and qualitative assessments. Ablation studies further validate the impact of our architectural choices, confirming the effectiveness of the innovative structures implemented within the network. Our research paves the way for more effective and efficient methods in pavement maintenance and monitoring, leveraging advanced techniques for practical, real-world applications in pavement management.

The conventional surface reflection method has been widely used to measure the asphalt pavement layer dielectric constant using ground-penetrating radar (GPR). This method may be inaccurate for in-service pavement thickness estimation with dielectric constant variation through the depth, which could be addressed using the extended common mid-point method (XCMP) with air-coupled GPR antennas. However, the factors affecting the XCMP method on thickness prediction accuracy haven’t been studied. Manual acquisition of key factors is required, which hinders its real-time applications. This study investigates the affecting factors and develops a modified XCMP method to allow automatic thickness prediction of in-service asphalt pavement with non-uniform dielectric properties through depth. A sensitivity analysis was performed, necessitating the accurate estimation of time of flights (TOFs) from antenna pairs. A modified XCMP method based on edge detection was proposed to allow real-time TOFs estimation, then dielectric constant and thickness predictions. Field tests using a multi-channel GPR system were performed for validation. Both the surface reflection and XCMP setups were conducted. Results show that the modified XCMP method is recommended with a mean prediction error of 1.86%, which is more accurate than the surface reflection method (5.73%).

Daily road maintenance is essential to road safety and serviceability, particularly to highways. In daily road maintenance inspection tasks, the objectives include a variety of targets, such as pavement cracks, foreign objects, guardrail damages etc. There is a lack of rapid detection methods that allow for the uniform identification of multiple targets for road maintenance. This paper proposes an automatic multi-object detection method for road daily maintenance inspection assisted by unmanned aerial vehicles (UAV). A dataset of multiple roadway anomalies (UAVROAD) for daily road maintenance needs was constructed. UM-YOLO, an improved algorithm based on the YOLOv8n algorithm was created to better extract the features of multiple targets, as well as fuse features and reduce computation while maintaining accuracy. The improvements include adding EMA (Efficient multiscale Attention Module) in the C2f module in the backbone, employing Bi-FPN fusion mechanism in the neck and using GSConv, a lightweight convolutional network, for the convolution operation. Compared with the YOLOv8n, the proposed UM-YOLO improved mean average precision(mAP) by 4.6% and reduced the model computation by 14%.

Neural network-based models have been implemented to predict various health indicators of asphalt pavement using pavement historical detection data. Unfortunately, their accuracy and reliability are not acceptable owing to their shallow architecture. To solve the issue, this study proposed an improved Transformer network to predict asphalt pavement health, called the Transformer with forward and reversed time series (Transformer FRTS). In terms of the input data, Transformer FRTS uses a new data form, so-called the random difference time series, to reduce the time dependency of the network prediction. In terms of the network architecture, the proposed network uses its encoder and decoder to obtain the data association from the forward and reverse time series. In addition, Transformer FRTS uses a post-processing decision criterion to improve the accuracy and reliability of prediction. The numerical experiment using the detection data from RIOHTrack full-scale track demonstrates that the proposed network has state-of-the-practice performance in asphalt pavement health prediction.

Image-based intelligent detection of road cracks with high accuracy and efficiency is vital to the overall condition assessment of the pavement. However, significant problems of continuous cracks interruption and background discrete noise misidentification are frequently observed in current semantic segmentation of pavement cracks, which mainly caused by traditional segmentation convolutional neural networks. This paper proposes a skip-level round-trip sampling block structure with the implementation of convolutional neural networks, thereby constructed a novel pixel level semantic segmentation network called CrackW-Net. After that, two datasets, including the widely recognized Crack500 dataset and a self-built dataset, were used to train two versions CrackW-Net, FCN, U-Net and ResU-Net. Meanwhile, comparative experiments are conducted among all these network models for crack detection. Results show that CrackW-Net without residual block performs the best in the task of pavement crack segmentation.

In transport sector, improving the quality of road construction plays a crucial role in developing a more sustainable and cost-effective infrastructure network. Intelligent Compaction (IC) has been actively studied as a novel and promising technology for quality monitoring and control in asphalt road construction. However, a successful application of IC in pavement construction is still bothered by its accuracy and stability in quality evaluation. This paper proposed AICV– Acceleration Intelligent Compaction Value, a new evaluation index for harmonic intelligent compaction quality evaluation and with a higher accuracy than the frequently used index CMV– Compaction Measurement Value. In order to optimize the evaluation of the compaction quality uniformity of asphalt pavement, the$3\sigma $of Normal Distribution is adopted to estimate the compaction uniformity. The statistical analyses of the field measurement data show that, both CMV and AICV are applicable to evaluate the uniformity of compaction quality. However, the dispersion degree of AICV is far less than that of CMV, indicating that AICV is more stable as an index for compaction quality evaluation. By using spatial statistics, the spatial correlation distance of the compaction quality is further obtained, which gives to an assessment of the influence range of intelligent roller. Overall, the study provides a basis for improving the quality control of asphalt pavement construction by means of more delicate monitoring of the compaction process.