Pavement Defects Research Articles

MGSCA: Multiscale Gradient Selection-Based Coupled-Hypersphere Adaptation for Unsupervised Pavement Defect Detection

MGSCA: Multiscale Gradient Selection-Based Coupled-Hypersphere Adaptation for Unsupervised Pavement Defect Detection

Machine Learning for Prediction of the International Roughness Index on Flexible Pavements: A Review, Challenges, and Future Directions

Timely maintenance of road pavements is crucial to ensure optimal performance. The accurate prediction of trends in pavement defects enables more efficient allocation of funds, leading to a safer, higher-quality road network. This article systematically reviews machine learning (ML) models for predicting the international roughness index (IRI), specifically focusing on flexible pavements, offering a comprehensive synthesis of the state-of-the-art. The study’s objective was to assess the effectiveness of various ML techniques in predicting IRI for flexible pavements. Among the evaluated ML models, tree ensembles and boosted trees are identified as the most effective, particularly in managing data related to traffic, pavement structure, and climatic conditions, which are essential for training these models. Our analysis reveals that traffic data are present in 89% of the studies, while pavement structure and climatic factors are featured in 78%. However, maintenance and rehabilitation history appears less frequently, included in 33% of the studies. This research underscores the need for high-quality, standardized datasets, and highlights the importance of model interpretability and computational efficiency. Addressing data consistency, model interpretability, and replicability across studies are crucial for leveraging ML’s full potential in fine-tuning IRI predictions. Future research directions include developing more interpretable, computationally efficient, and less complex models to maximize the impact of this research field in road infrastructure management.

CNN-based pavement defects detection using grey and depth images

This paper introduces a method for detecting pavement defects based on convolutional neural networks. First, grey and depth image data were acquired using a 3D pavement information collection system, followed by pre-processing and labelling of the data. Subsequently, two network structures were developed to accommodate the image data characteristics: classic U-shaped and double-headed structures. Attention modules were integrated into the models to enhance the accuracy of defect detection. Finally, a quantitative analysis of four types of pavement defects was conducted. The numerical evaluation results demonstrated that training the network with a combination of grey and depth images significantly improves the detection accuracy, resulting in a 10% enhancement in mean intersection over union (MIoU). The proposed model attained a global pixel accuracy (GPA) of 97.36% and an MIoU of 80.28%. The proposed network model was found to have an increased focus on the pavement defect areas, making it highly effective.

A review into studies related to the effect of the pavement surface condition on traffic safety: A scientometric analysis

Traffic crashes are one of the main reasons for the death of many people and the loss of property Therefore, it is important to conduct research and studies to reduce the risk of accidents and identify the causes that lead to their occurrence. The condition of the pavement is one of the main factors that lead to accidents, as several studies have been presented that show the impact of pavement defects such as rutting, roughness, and skid resistance on traffic safety. This research was conducted to find out and compile the most important research papers using the web of science (WOS) and then analyzing the data using the VOSviewer program, and to know the most countries and journals published on the related pavement condition of road crashes, as well as to know the authors and their cooperation. In addition to knowing the keywords that help researchers research this topic. In order to benefit from these studies in knowing the causes of accidents, analyzing and treating them, and improving the performance of roads using modern analysis and maintenance methods that ensure traffic safety.

Evaluation of the Pavement Distress and its Impact on the Sustainability of the Traffic Operation for Selected Roads in Al-Diwaniyah City

Roads and transportation infrastructure are critical assets for supporting political stability as well as economic and sustainable expansion in developing countries. Yet, pavement maintenance backlogs and the high capital expenditures of road rehabilitation need the adoption of pavement evaluation methods to ensure the best value for the investment. road maintenance is considered one of the main factors for preserving the road structurally and functionally, and before carrying out maintenance, pavement defects must be identified by evaluating the condition of the pavement to perform the appropriate maintenance. In this study, the Pavement Condition Index (PCI) method was relied upon to evaluate roads in Al-Diwaniyah City. Where several data were collected for the study, including the number of crashes for each region, the number of lanes, street width, the presence of a median, and the pavement condition evaluated for the regions with the highest crash rate, which are Al-Askari District with 707 crashes, Al-Urouba District with 379 crashes, Al-Wahda District with 271 crashes, Al-Jamhouri District with 235 crashes, and Al-Jazaer District with 212 crashes. This study is considered the first study to evaluate roads in Al-Diwaniyah city. The results of the study showed that the condition of the roads ranged from poor to satisfactory and there is no road reach to serious and failed. The results showed that the worst road is Al-Askari District, which lacks a lot of maintenance, at a rate of PCI 52.1. It is followed by Al-Wahda District with a PCI rate of 57.6, Al-Jamhouri District with a PCI rate of 67.2, Al-Urouba District with a PCI rate of 70.7, and Al-Jazaer District with a PCI rate of 73.7. The deterioration of the pavement surface condition affects the rash rate, and it is more dangerous when it is correlated to other road factors. Additionally, it was noted that the most prominent defects in the pavement surface of the five evaluated roads are: potholes, rutting, longitudinal and transverse cracks, corrugation, patching, raveling, depressions, and alligator cracks. The study recommended conducting more research on evaluating more roads in Al-Diwaniyah City on an annual basis and conducting periodic maintenance in order to improve the status of this city, which suffers from neglect in all respects, measures must be taken to enhance the environment and economy, sustain the streets, reduce maintenance costs, ensure traffic safety, and protect the lives of citizens.

Lightweight Model for Pavement Defect Detection Based on Improved YOLOv7.

Existing pavement defect detection models face challenges in balancing detection accuracy and speed while being constrained by large parameter sizes, hindering deployment on edge terminal devices with limited computing resources. To address these issues, this paper proposes a lightweight pavement defect detection model based on an improved YOLOv7 architecture. The model introduces four key enhancements: first, the incorporation of the SPPCSPC_Group grouped space pyramid pooling module to reduce the parameter load and computational complexity; second, the utilization of the K-means clustering algorithm for generating anchors, accelerating model convergence; third, the integration of the Ghost Conv module, enhancing feature extraction while minimizing the parameters and calculations; fourth, introduction of the CBAM convolution module to enrich the semantic information in the last layer of the backbone network. The experimental results demonstrate that the improved model achieved an average accuracy of 91%, and the accuracy in detecting broken plates and repaired models increased by 9% and 8%, respectively, compared to the original model. Moreover, the improved model exhibited reductions of 14.4% and 29.3% in the calculations and parameters, respectively, and a 29.1% decrease in the model size, resulting in an impressive 80 FPS (frames per second). The enhanced YOLOv7 successfully balances parameter reduction and computation while maintaining high accuracy, making it a more suitable choice for pavement defect detection compared with other algorithms.

Evaluation of Pavement Surface Distress Using Image Processing and Artificial Neural Network

Abstract Pavement evaluation helps assess the structural and functional condition of roads. Traditional pavement evaluation methods cannot efficiently and accurately assess the state of road conditions at the network level. Though the pavement evaluation techniques based on high-resolution cameras and laser sensors assess the road conditions efficiently at traffic speeds, there are a few limitations in respect of automotive detection and quantification of pavement surface distress. Though artificial intelligence is a broad area, machine learning applications in highway engineering were proven to be successful. This paper presents the automated pavement distress classification using a convolutional neural network (NN) from the Keras library. VGG-16, the deep convolutional NN (DCNN) model, was deployed by necessary modification to get the desired output. The model is trained on a big data set of images with a wide range of pavement defects and irregularities. A DCNN classifier trained with an “Adam” optimizer was used to change the features of the NN to minimize the loss. As an output, the model classifies the pavement surface distresses as alligator cracks, longitudinal cracks, transverse cracks, pothole, and no crack portion of the pavement with maximum accuracy using a DCNN classifier.

Using supervised machine learning algorithms in pavement degradation monitoring

• Dynamic pavement monitoring system using vibration-based method. • Noise cancelling and vibration data smoothing. • Degradation detection and classification using machine learning. Pavement monitoring plays a vital role in maintaining sustainable road network conditions and provides road users with satisfactory comfort riding. Regular and routine monitoring provide clear information on road conditions and the level of damage to pavement surfaces. In this study, a vibration-based method was used as a monitoring technique to evaluate the pavement surface conditions on local roads. Pre-processing techniques were used to cancel noise and prepare the data for feature extraction and prediction. Two multi-classification Machine Learning (ML) models were used, including Random Forest (RF) and Decision Tree (DT), for the automated classification and detection of different types of pavement distresses. In addition, a Support Vector Machine (SVM) technique was used to develop a binary ML model for the same classification and detection purposes. The results showed that the developed ML models provide high accuracy in predicting the road degradation classification with about 93% accuracy using the RF and 90% accuracy using the DT. Using the SVM model, the overall average accuracy of detection and classification of pavement defects was about 96%.

Semantic segmentation of pavement cracks based on an improved U-Net

Cracks are one of the types of pavement defects that can affect the safety and quality of roads, so identifying such defects is an important part of road maintenance. In this paper, based on the U-Net coding-decoding structure, an ECA channel attention module is added to the coding stage, thus improving feature extraction without introducing too many extra parameters and computational effort. The insertion of the FCNhead decoding dock in the decoding stage can improve the performance of the model on semantic segmentation tasks while maintaining its efficiency and interpretability, thus better meeting the needs of practical applications. Experimental results on the CFD dataset and the crack500 dataset show that the algorithm improves the accuracy of crack detection and has better robustness.

Impact of Pavement Defects on Traffic Operational Performance

The present research investigates the influence of pavement distresses on traffic operational performance in terms of speed and capacity. The pavement distresses are expressed by the pavement condition index (PCI). Three roads of multi-lane highways located on the road network of Kafrelsheikh governorate, Egypt, were chosen as a case study. Road geometry, pavement condition characteristics, and traffic volume were collected manually from 23 sections for field data surveys. For speed measurements, spot speed data were gathered using the stopwatch method under the free-flow conditions, and the percentile speeds (V50% and V85%) have been calculated for different vehicles classes. Traffic capacity was determined via extrapolation of the quadratic function. Regression analysis models were created to express the relationships of both the percentile speeds and traffic capacity as dependent variables, and three independent variables (PCI, lane, and shoulder widths). These empirical mathematical relations help in developing road design and maintenance works on multi-lane highways. The main conclusion of the study demonstrated that traffic operational performance was more significantly affected by pavement distresses than road geometry. Moreover, the results indicated that pavement distresses have a significant impact on light vehicles more than heavy ones.

Multi-scale triple-attention network for pixelwise crack segmentation

Currently, intelligent crack detection is of great value for the maintenance of infrastructure, of which the most significant kind in China is roads. For pavement defects, the pavement can be repaired and maintained in a timely manner with an accurate defect detection task, which significantly reduces the occurrence of hazards. However, the detection of pavement defects remains a great challenge owing to many difficulties, for example, complex backgrounds, microdefects, various defect shapes and sizes, class imbalance issues, etc. Recently, deep learning has demonstrated its superior performance on pixelwise image segmentation, but some issues still exist on demanding pixelwise image segmentation, for instance, limited receptive field, insufficiency processing of local features, information loss issue generated by pooling operations, etc. Based on all of the above issues, a multiscale triple-attention network, named MST-Net, is proposed for end-to-end pixelwise crack detection. First, a multiscale input strategy is applied to the proposed segmentation network to capture more context information. Meanwhile, it can capably reduce the effect of the information loss issue generated by pooling operations. Second, to realize effective feature representation of local features, an additive attention fusion (AAF) block is proposed to guide feature learning to capture both global and local contexts. In addition, faced with the crack detection task with class imbalance issues, a triple attention (TA) block is proposed to detect spatial, channel and pixel attention information to suppress the background and useless information, which is conducive to the characterization of microcracks. Finally, aiming at the limited receptive field, a multiscale feature aggregation unit is proposed for feature fusion to increase the detection ability of multiscale defects. To better guide network training, a deep supervision mechanism is also introduced to speed up the convergence of the proposed segmentation model and improve the performance of defect segmentation. The related evaluation and detection experiments are carried out on three public datasets on crack segmentation, and the comparison experiments with the mainstream segmentation models show that the proposed segmentation network achieves excellent performance on pixelwise crack detection.

Normal-Knowledge-Based Pavement Defect Segmentation Using Relevance-Aware and Cross-Reasoning Mechanisms

Automatic pavement defect segmentation is a big challenge because of the class diversity and extremely random distribution of defects. Most existing approaches focus on supervised strategies to achieve decent performance. Due to the difficulty of getting massive densely annotated samples and the limited prior knowledge of potential defects, these methods have significant bottlenecks in the actual pavement settings. This paper proposes a relevance-aware and cross-reasoning network (RCN) for anomaly segmentation of pavement defects, which can segment defects using merely non-defective images for training. A relevance-aware transformer-based encoder is first devised to model intrinsic interdependencies across local features, thus improving representations of complex non-defective images. Next, a dual decoder strategy is proposed to remap the encoder-generated latent dependencies at the local semantic and global detailed levels, respectively. Specifically, a cross-reasoning refinement module is built in the local decoder to reason the cross-relationship between spatial and channel dimensions. Finally, a context-aware abnormal distillation measurement is developed to evaluate the semantic reconstruction deviations during the inference. Under the guidance of semantic affinity, this measurement allows our model to highlight defective areas adaptively. Extensive experimental results on four datasets indicate that RCN outperforms other leading anomaly segmentation methods.

A 5G Cloud Platform and Machine Learning-Based Mobile Automatic Recognition of Transportation Infrastructure Objects

Crack recognition is important in periodic pavement inspection and maintenance. The wide application of image recognition technology in daily inspection and maintenance makes the health monitoring of asphalt pavement defects more effective, both intelligently and sustainably. In this study, a mobile automatic system integrating fifth-generation wireless communication technology (5G), cloud computing, and artificial intelligence (AI) was proposed for transportation infrastructure object recognition. The original dataset contained 344 images of pavement defects, including longitudinal cracks, transverse cracks, alligator cracks, and broken road markings. Three lightweight algorithms for automatic pavement crack identification were used and compared, including MobileNetV2, ShuffleNetV2, and Res-Net50 networks, respectively. The results showed that the model based on ShuffieNetV2 achieved the best overall predictive accuracy (ACC = 95.52 percent). A mobile automatic monitoring system based on the cloud platform and Android framework was then established. With the help of 5G technology, the cloud-network-terminal’ interconnection can be achieved to provide fast and stable information transmission between transportation infrastructure and road users. The proposed system provides an engineering reference for the transportation infrastructure inspection and maintenance using the 5G communication technology.

Comparison of histogram-based gradient boosting classification machine, random Forest, and deep convolutional neural network for pavement raveling severity classification

Raveling is a widely encountered defect found in asphalt pavements. Raveling deteriorates riding safety and accelerates the development of other pavement defects. Therefore, timely and accurate classification of raveling severity is required to establish cost-effective pavement maintenance plans. This paper compares the performances of the Histogram-Based Gradient Boosting Classification Machine (HGBCM), Random Forest (RF), and Deep Convolutional Neural Network (DCNN) in recognizing raveling severity. Texture analysis based on the Gray-Level Co-Occurrence Matrix and statistical measurements of color channels is used for feature extraction. An image dataset including 3600 samples and three categories (non-raveling, minor raveling, and severe raveling) has been collected to train and verify the proposed methods. Experimental results supported by statistical hypothesis tests point out that HGBCM has achieved an outstanding performance with an accuracy rate > 0.96, F1 score > 0.94, and Cohen's Kappa coefficient > 0.92 for all class labels.

Approximation of Pavement Key Performance Indicators from Crowdsourced Data

The feasibility of Pavement KPI approximation from crowdsourced data is examined. A dedicated smartphone application for the collection of positioning and accelerometer data as well as regular manual driver inputs is used. Smartphone application data is correlated with KPIs determined using pavement defect and International Roughness Index data collected with a Laser Crack Measuring System. Statistical and machine learning models are developed for the approximation of KPIs from smartphone data, and then the resulting models are used to predict pavement KPIs beyond the training dataset. The performance of alternative models is examined and the cost savings of such an approach in the deployment and running of a Pavement Asset Management System.

Investigating the Risk Factors on Crash Severity for Selected Risky Roads in Al-Diwaniyah City by Utilizing a Binary Probit Model

Traffic crashes are one of the main reasons for the deaths of many people and the loss of property. Road safety is a crucial aspect of transportation that aims to prevent crashes and injuries on the road, and several contributing factors affect it. In this study, the binary probit model using the N-Logit software was applied to crash-related data to examine the contribution of several variables to severe crash outcomes in Al-Diwaniyah City. Crash severity (the dependent variable) in this study is a dichotomous variable with two categories, severe and non-severe. Because of the binary nature of this dependent variable, a binary probit model was found suitable. Out of 37 independent variables obtained from Al-Diwaniyah Hospital traffic crash reports between 2014 and 2021 and from fieldwork to evaluate the pavement surface condition using the pavement condition index (PCI), six were statistically significantly associated with severe crashes. These factors include driver age, spring and summer seasons, pavement surface conditions, pedestrian collisions, and multi-vehicle crashes. Some proposals were also recommended to reduce traffic crashes, such as using median barriers to regulate pedestrian crossing, managing a proper number of lanes in the roads to avoid congestion due to a large number of vehicles, assessing the condition of the pavement surface annually at least, to identify pavement defects and conduct appropriate maintenance. Therefore, governments and transportation agencies must prioritize regular pavement condition evaluation as part of their maintenance programs for roads and highways. In conclusion, road safety is a complex issue that requires a multi-faceted approach involving various stakeholders such as government agencies, law enforcement agencies, vehicle manufacturers, drivers, pedestrians, and other road users.

MS-FPN-Based Pavement Defect Identification Algorithm

MS-FPN-Based Pavement Defect Identification Algorithm

Multiscale Attention Networks for Pavement Defect Detection

Multiscale Attention Networks for Pavement Defect Detection

Defects in the road surface of bridges, which affect the nature of the movement of motor vehicles

The article conducts a study to identify typical pavement defects and their geometric parameters that affect the nature of vehicle traffic and lead to excessive dynamic loads on road bridges not provided for by the design. Various research methods are used, including measurement of vehicle movement parameters, vehicle movement theory, diagnostics of bridge deck defects, and analysis of the impact of defects on automotive equipment. Accordingly, the results obtained in the article are based on the combination and analysis of information from specialists in road construction and automotive engineering. The results obtained can be used in mathematical modeling of the movement of vehicles over road bridges with defects and damage to the road surface.

Road Maintenance Practices in Nepal: A Case Study of Malekhu – Mugling Road Section

Road degradation is a major issue in Nepal due to main focus only on backlog maintenance and upgradation which evidently is an unsuccessful maintenance investment strategy worldwide. Considering global success of PBMC in time-cost-quality spectrum, it has been implemented in Nepal since 2003 but without success. Thus, ADB in 2016 revised PBM specifications and re-executed it, whose performance potential must be accessed with respect to current SMDP practice. The current study analyzes and compares effectiveness of SMDP and PBMC in terms of cost and quality for Malekhu- Mugling road section, and suggests a suitable maintenance practice in the current scenario of Nepal. This study reveals that strengthening SMDP practice shall be focused on before releasing a long term PBMC. The study concludes that the most suitable practice in terms of cost quality optimization is a strengthened SMDP as per PBMC standard whereby length worker perform PBMC standard routine, pavement and intervention based repairs due to their reliability during emergency. Then, one year contract shall be released incorporating two recurrent (prior and post monsoon) and a specific contract where contractors essentially follow performance based maintenance of roadside structures, traffic safety, pavement defects if any, and any emergency maintenance during monsoon. It is recommended that PBMC model shall be developed gradually from one year contract to multiyear contracts since there may be high chances of claims and conflicts in four year long PBMC.