Pavement Performance Prediction Research Articles

Enhancing pavement performance prediction with traffic speed deflection data

ABSTRACT Pavement performance modelling is an essential part of pavement management. Past research has shown that a structural indicator can be included in pavement performance models to capture differences in performance between weak and strong pavements. Weak pavements are expected to deteriorate more rapidly than strong pavements, as consistently shown in past research. The work presented in this paper uses the Traffic Speed Deflectometer (TSD) to calculate multiple structural indexes and compares their ability to differentiate between poor and well performing pavements. Historical condition data from the US National Parks pavement network is used in this assessment. Two approaches for calculating the subgrade resilient modulus and two approaches for calculating the effective structural number from the TSD results are compared. The temperature corrected centre deflection, surface curvature index and ratio of effective to required structural number are all calculated from the TSD data. Performance models are developed for two condition indexes and the International Roughness Index (IRI). In all cases, the addition of the structural indexes improved the quality of predictions, and the IRI prediction showed the most improvement from including the indexes. Ultimately, it was shown that the TSD testing results can be used to inform the pavement management process.

Overview and Discussion of Pavement Performance Prediction Techniques for Maintenance and Rehabilitation Decision-Making

Overview and Discussion of Pavement Performance Prediction Techniques for Maintenance and Rehabilitation Decision-Making

Revolutionising Egyptian pavement design: a comprehensive E* database and advanced modeling approach for contextually informed performance predictions

ABSTRACT This research assesses the dynamic modulus (E*) as a pivotal rheological attribute for characterizing viscoelastic behaviour in various indigenous asphalt mixtures across diverse scenarios. An exhaustive investigation involving 32 laboratory-designed asphalt mixes and two field mixes explores the influence of traffic loading patterns, air voids, binder sources, and climatic conditions on E*. Utilizing Superpave gyratory compaction, 68 E*-specimens are compacted and subjected to E*-testing, leading to master curve development. Evaluation of NCHRP 1-37A and NCHRP 1-40D Witczak models demonstrates their fair to excellent accuracy in predicting E* for Egyptian mixes. The NCHRP 1-40D model stands out with an R² of 0.94, showcasing superior accuracy and minimal bias. This study integrates AASHTOW are Pavement ME Design (PMED) to analyse pavement sections, revealing the impact of design factors, characterized by measured E*, on predicted pavement performance, encompassing asphalt concrete rutting, alligator cracking, longitudinal cracking, and terminal international roughness index. The investigation highlights the marginal impact of substituting predicted E* for measured E* on AASHTOWare pavement performance indicators for Egyptian mixes under default binder characteristic values. This research contributes valuable insights into the interplay of E* and pavement performance, facilitating wellinformed decisions in pavement design and analysis across diverse conditions in the Egyptian context.

Novel Instance-Based Transfer Learning for Asphalt Pavement Performance Prediction

The deep learning method has been widely used in the engineering field. The availability of the training dataset is one of the most important limitations of the deep learning method. Accurate prediction of pavement performance plays a vital role in road preventive maintenance (PM) and decision-making. Pavement performance prediction based on deep learning has been widely used around the world for its accuracy, robustness, and automation. However, most of the countries in the world have not built their pavement performance historical database, which prevents preventive maintenance using the deep learning method. This study presents an innovative particle swarm optimization (PSO) algorithm-enhanced two-stage TrAdaBoost.R2 transfer learning algorithm, which could significantly increase the pavement performance prediction database. The Long-Term Pavement Performance (LTPP) database is used as the source domain data, and one of the highways in China is chosen as the target domain to predict pavement performance. The results show that the proposed PSO-Two-stage TrAdaBoost.R2 model has the highest accuracy compared with AdaBoost.R2 model and traditional regression decision tree model. The validation case study shows significant consistency between the predicted International Roughness Index (IRI) and the whole-year measurement data with an R2 of 0.7. This study demonstrates the great potential of the innovative instance-based transfer learning method in pavement performance prediction of a region’s lack of data. This study also contributes to other engineering fields that could greatly increase the universality of deep learning.

Predictive models for flexible pavement fatigue cracking based on machine learning

Pavement performance prediction is crucial for ensuring the longevity and safety of road networks. In our extensive study, we employ a diverse array of techniques to enhance fatigue performance models in flexible pavements. The methodology begins with Random Forest feature selection, identifying the top 15 critical variables that significantly impact pavement performance. These variables form the basis for subsequent model development. Our investigation into model performance indicates the superiority of advanced machine learning methods such as Regression Trees (RT), Gaussian Process Regression (GPR), Support Vector Machines (SVM), Ensemble Trees (ET), and Artificial Neural Networks (ANN) over traditional linear regression methods. This consistent outperformance underscores their potential to reshape forecasting accuracy. Through extensive model optimization, we reveal robust performance across both complete and selected feature sets, emphasizing the importance of meticulous feature selection in enhancing forecast accuracy. The accuracy of our best optimized machine learning model is highlighted by its Performance Measurement metrics: RMSE of 22.416, MSE of 502.46, R-squared of 0.80848, and MAE of 8.9958. Additionally, comparative analysis with previous empirical models demonstrates that our best optimized machine learning model outperforms existing empirical models. This work underscores the significance of feature curation in pavement performance prediction, highlighting the potential of sophisticated modeling methodologies. Embracing cutting-edge technologies facilitates data-driven decisions, ultimately contributing to the development of more robust road networks, enhancing safety, and prolonging lifespan.

Enhancing rutting depth prediction in asphalt pavements: A synergistic approach of extreme gradient boosting and snake optimization

This study investigates the development of pavement performance models, focusing on the predictive analysis of pavement distress using ensemble machine learning methods. We employ the Research Institute of Highway Ministry of Transport track dataset to construct predictive models for pavement distress, comparing various methodologies including Random Forest (RF) regression, Extreme Gradient Boosting (XGB) machines, Snake Optimizer-Random Forest (SO-RF), and Snake Optimizer-Extreme Gradient Boosting (SO-XGB). The approach is conducted by multiple experiments and a comprehensive comparative analysis leveraging the repeated K-Fold cross-validation method to ensure the robust assessment for these models. Our findings show that the SO-XGB method is particularly effective in predicting rutting depth. The strength of the SO-XGB model lies in its excellent performance in several performance metrics, such as coefficient of determination (R2), the mean absolute error (MAE), root mean square error (RMSE), the mean absolute percentage error (MAPE), performance index (PI), index of agreement (IA), variance accounted for (VAF) and objective function (OBJ). These results not only illustrate the model’s accuracy but also the practical applicability of SO-XGB model in automatically forecasting pavement damage. Thus, the SO-XGB method stands out as a direct, powerful, and efficient tool in the field of pavement performance prediction.

Numerical analysis of field aging in asphalt pavements via coupling of diffusion process and kinetics aging model

The asphalt aging model is a crucial module in the Mechanistic-Empirical (ME) pavement design programme for asphalt pavement performance prediction. In this study, an analytical aging model that couples diffusion and kinetics aging model (CDKAM) is improved to analyze and predict the field aging of asphalt pavements. The complex modulus (E*) of field cores was used as the aging property to illustrate the derivation of explicit expressions of time-averaged thermal source term ( TST ¯ ( t ) ) and time-averaged diffusion coefficient ( D ¯ ( t ) ). The value of the material-related parameter (M) in TST ¯ ( t ) model differs among different mixtures, which could assess the mixture’s susceptibility to thermal oxidation. The CDKAM with TST ¯ ( t ) and D ¯ ( t ) was used to predict E* of field aging for asphalt pavements in Texas after 22-month service life. The calibrated CDKAM could reasonably quantify the aging extent over time and the aging gradient over depth for studied asphalt pavements.

Quantifying the Effects of Material Input Levels on Jointed Plain Concrete Pavement (JPCP) Performance and Slab Thickness

The mechanistic-empirical pavement design guide (MEPDG) is a commonly accepted design principles guide that aids in jointed plain concrete pavement (JPCP) design and performance analysis. The MEPDG uses three different design parameter input levels, referred to as level one, level two, and level three, providing increasing confidence in the analysis at the lower numbered levels, which use more locally relevant (level two) or project-specific (level one) data. The state-of-the-art pavement ME software (version 2.6.2) uses MEPDG design principles to predict pavement performance. The three performance indicators for JPCP systems (international roughness index (IRI), joint faulting, and transverse cracking) experience significant changes when simulating under a different input level. The IRI and faulting indicator changed by 78 percent when using inputs varying from level one to level three, with the cracking indicator change being more severe at 87 percent. To accommodate the change in performance indicator values between input level one and input level three, increasing the concrete slab thickness is necessary to achieve comparable pavement performance. An increase in the Portland cement concrete (PCC) layer from one inch to two inches is required when input level three simulations are performed, demonstrating the economic and sustainability benefits of using project-specific level one inputs. Understanding the impact of simulation input levels will help to meet design and sustainability goals and improve the lifecycle performance of JPCP systems.

Development of a Mechanistic-Empirical-Based Highway Cost Allocation Model for Flexible Pavements

Construction, operation, and maintenance of a pavement network requires funding, partially sourced from road user taxes. Recent studies showed that lightweight vehicles are typically taxed higher compared with heavy trucks that damage the roads the most. To facilitate the equity and fairness of the allocated costs to different vehicles in the United States (U.S.), Highway Cost Allocation Studies (HCAS) were performed using various pavement performance prediction models. Reviewed literature showed the lack of mechanistic-empirical (ME)-based HCAS models for the flexible pavement network. In this study, a national-level ME-based HCAS model was developed, and the damage shares of different vehicle classes have been estimated for 67,583 pavement sections in the U.S. Highway Performance Monitoring System (HPMS) database. The proposed HCAS model was compared with the existing Federal Highway Administration (FHWA) HCAS model (i.e., National Pavement Cost Model [NAPCOM]). The analysis of the traffic data showed that two-axle single-unit trucks (SU2) and tractor-semitrailers with two tandem and one single axle (CS5T) were the most frequent users of the pavement network. The results showed that the damage share of SU2 is dominant in minor roadways, while the damage share of the heavier vehicles in the CS5T class is dominant in major arterials and interstates. In addition, it was found that, although the geographical location and environmental condition of the pavement section affects the magnitude of the pavement distresses, the distribution of the damage shares remains almost the same. This can be attributed to the similarities in the traffic data, for example, vehicle class distribution and axle load spectra.

Research and comparison of pavement performance prediction based on neural networks and fusion transformer architecture

<abstract> <p>The decision-making process for pavement maintenance from a scientific perspective is based on accurate predictions of pavement performance. To improve the rationality of pavement performance indicators, comprehensive consideration of various influencing factors is necessary. To this end, four typical pavement performance indicators (i.e., Rutting Depth, International Roughness Index, Longitudinal Cracking, and Alligator Cracking) were predicted using the Long Term Pavement Performance (LTPP) database. Two types of data, i.e., local input variables and global input variables, were selected, and S-ANN and L-ANN models were constructed using a fully connected neural network. A comparative analysis of the predictive outcomes reveals the superior optimization of the L-ANN model. Subsequently, by incorporating structures such as self-attention mechanism, a novel predictive approach based on the Transformer architecture was proposed. The objective is to devise a more accurate predictive methodology for pavement performance indices, with the goal of guiding pavement maintenance and management efforts. Experimental results indicate that, through comparative analysis of three quantitative evaluation metrics (root mean square error, mean absolute error, coefficient of determination), along with visual scatter plots, the predictive model employing the fused Transformer architecture demonstrates higher robustness and accuracy within the domain of pavement performance prediction when compared to the L-ANN model. This outcome substantiates the efficacy and superiority of the model in terms of predictive performance, establishing it as a reliable tool for accurately reflecting the evolution of asphalt pavement performance. Furthermore, it furnishes a theoretical reference for determining optimal preventive maintenance timing for pavements.</p> </abstract>

Improved empirical convection heat transfer coefficient model to predict flexible pavement layer temperatures

Temperatures in asphalt layers of a flexible pavement significantly influence its performance, and its accurate prediction is vital for effective life-cycle design that models asphalt as a viscoelastic material. In 2023, Rajapaksha et al. observed that convective heat transfer plays a significant role in temperature prediction, and this study focused on developing a more accurate model for Convection Heat Transfer Coefficient (CHTC) to achieve that using a five-step approach. The cyclical nature of temperatures measured at 5-minute time steps for one year prompted a model based on Fourier analysis. Step 1 conducted a parametric study of CHTC that compared predicted and measured temperatures to backcalculate an optimum value for each time step. Step 2 used these values as the basis to develop monthly Fourier models, but they were not sufficiently accurate. When key weather parameters were added to it in Step 3, the resulting monthly empirical models developed using stepwise regression analysis improved to an acceptable level. In Step 4, the monthly CHTC regression models were used to predict pavement temperatures for the whole year that were compared with measured values from a field test section. An error analysis performed in Step 5 revealed that the RMSE of prediction error for the year was 2.178 °C which was very close to the benchmark set at the beginning of this work. In addition, the median error was estimated at −0.09 °C showing that the proposed CHTC model is sufficiently accurate to be used in pavement performance prediction.

Prediction model of asphalt pavement functional and structural performance using PSO-BPNN algorithm

The rapid expansion of expressway mileage has resulted in a significant maintenance workload. Accurately predicting pavement performance is an effective approach to cost-saving, maintaining the pavement in optimal condition, and prolonging its service life. However, the current asphalt pavement performance prediction model demonstrates low accuracy and fails to fully capture the actual changes in pavement performance under varying traffic and environmental conditions. In this study, we propose an enhanced back propagation neural network (BPNN) prediction model based on the particle swarm optimization (PSO) algorithm for predicting six performance parameters of asphalt pavement, encompassing both functional and structural aspects. The influential factors considered as inputs for our model include surface layer thickness, traffic load, and climate conditions, while the output is represented by specific pavement performance indices. To evaluate the effectiveness of our proposed model, three performance indices - root mean square error (RMSE), mean absolute percentage error (MAPE), and R-squared (R2) - were employed, leading us to obtain an optimized PSO-BPNN prediction model structure. We validated our results using road surface detection data from a Chinese highway spanning from 2016 to 2020, thereby demonstrating the feasibility of our PSO-BPNN prediction model. Our case study reveals that compared to the pre-optimized BPNN model as a whole, the optimized BPNN model achieved a higher R2 value by 17.43%, with an average R2 value exceeding 0.878 across all performance indices, additionally maintaining MAPE within 5%. These findings indicate that our optimized model exhibits faster convergence speed along with improved predictive capability and generalization effect. Consequently, this developed model can be effectively integrated into a pavement management system for generating accurate and timely maintenance plans.

Dynamic Modulus Prediction Validation for the AASHTOWare Pavement ME Design Implementation in Egypt

Dynamic Modulus, E* is a crucial property of the hot mix asphalt (HMA). For the AASHTOWare Pavement ME design, E* is an essential material input. E* can be measured in the laboratory or predicted using different models based on some fundamental properties of the HMA. The NCHRP 1-37A and NCHRP 1-40D prediction models are the two main models adopted by the AASHTOWare to predict the E* based on the HMA mixture volumetrics, gradation, and binder properties. The main objective of this research was to validate these two prediction models using local HMA mixes for the purpose of the regional application of the AASHTOWare Pavement ME design in Egypt. For this purpose, the E* values of ten locally plant-produced HMA mixes were measured in the laboratory. The two E* prediction models were then used to estimate the E* values for the same materials. Consequently, the performance of both models was studied by comparing the measured values to the estimated values. The results showed that the NCHRP 1-40D prediction model can satisfactorily predict the E* of the Egyptian HMA mixes with minimal bias and high accuracy. The model yielded an adjusted coefficient of determination (R2) of 0.86 based on 480 E* measurements. On the other hand, the NCHRP 1-37A prediction accuracy was not satisfactory, with very poor accuracy (Adjusted R2 = 0.18) and high bias. Afterwards, the effect of the predicted E* from the NCHRP 1-40D model on the AASHTOWare Pavement ME predicted pavement performance in terms of rutting, cracking, and roughness was further studied. Accordingly, twenty-four simulation runs for typical Egyptian design cases were conducted using, first, the laboratory measured E* values and, then, the NCHRP 1-40D predicted E* values. The results showed that the NCHRP 1-40D predictions had no significant effect on the pavement performance predicted by the AASHTOWare Pavement ME with R2 of the different pavement distresses ranged from 0.980, for the AC rutting, to 0.9996 for the International Roughness Index (IRI). Hence, the NCHRP 1-40D model can be used satisfactorily to predict E* for the Egyptian HMA mixes without compromising the structural pavement design.

Machine learning based pavement performance prediction for data-driven decision of asphalt pavement overlay

This study is to develop pavement performance models using support vector regression and ensemble machine learning methods for selection of pavement overlay strategy. Predictive models of pavement distresses were developed based on the Long-Term Pavement Performance (LTPP) data and compared using support vector machine, random forest regression, gradient boosting machine, and stacking ensemble. Gradient boosting machine was found to be a more effective method to establish predictive models of rut depth and International Roughness Index. Stacking ensemble and random forest regression would provide reliable prediction of alligator cracking. The models developed based on the clusters of climate and traffic parameters were found to be more effective. Based on the developed performance models, the effects of asphalt overlays on pavement distresses and service lives were investigated. When the overlay with recycled asphalt concrete (AC) was applied, the propagation of alligator cracking was faster compared to the overlay with virgin asphalt mixture. Milling before overlay tended to slow the increase of IRI but fasten the development of rut depth.

Predicting pavement performance using distress deterioration curves

Highway Authorities in the UK use Surface Condition Assessment for the National Network of Roads (SCANNER) in assessing and managing their road networks. This survey vehicle utilises laser measurements to detect and quantify most of the distress on the road surface, such as rutting, cracking and texture depth. It is however a data intensive and expensive approach since it is conducted annually. This study presents a simple method to predict pavement distress using previous SCANNER measurements. The previous measurements are used to develop Distress Deterioration Master Curves (DDMC) that relate distress deterioration rate with the severity of the distress. These curves can be used to predict future distress severity based on the current state without the need to provide further data such as pavement age or pavement material properties. To demonstrate the application of this method, a significant amount of SCANNER data covering around 400 km of class A roads in Nottinghamshire collected between 2014 and 2020 were analysed, and rutting, crack intensity, and texture depth were modelled in this study. DDMRs of these distress types were built based on data collected between 2014-2018, then 2020 data were used to validate the predictions. The results show that the developed method can be implemented in predicting surface distress of roads using previous measurements, which makes it a valuable addition tool for highway authorities subject to underfunding.

A machine learning-based approach to assess impacts of autonomous vehicles on pavement roughness.

Studies have been initiated to investigate the potential impact of connected and automated vehicles (CAVs) on transportation infrastructure. However, most existing research only focuses on the wandering patterns of CAVs. To bridge this gap, an apple-to-apple comparison is first performed to systematically reveal the behavioural differences between the human-driven vehicle (HDV) and CAV trajectory patterns for the first time, with the data collected fromthe camera-based next generation simulation dataset and autonomous driving co-simulation platform, CARLA and SUMO, respectively. A gradient boosting-based ensemble learning model for pavement performance (i.e. international roughness index) prediction is then developed with the input features including three driving pattern features, namely, lateral wandering deviation, longitudinal car-following distance and driving speed, plus 20 othercontext variables. A total of 1707 observations is extracted from the long-term pavement performance database for model training purposes. The result indicates that the trained model can accurately predict pavement deterioration and that CAV deteriorates pavement faster than HDV by 8.1% on average. According to the sensitivity analysis, CAV deployment will create a greater impact on the younger pavements, and the rate of pavement deterioration is found to be stable under light traffic, whereas it will increase under congested traffic. This article is part of the theme issue 'Artificial intelligence in failure analysis of transportation infrastructure and materials'.

Use of recurrent neural networks considering maintenance to predict urban road performance in Beijing, China.

A correct understanding of the pavement performance change law forms the premise of the scientific formulation of maintenance decisions. This paper aims to develop a predictive model taking into account the costs of different types of maintenance works that reflects the continuous true usage performance of the pavement. The model proposed in this study was trained on a dataset containing five-year maintenance workdata on urban roads in Beijing with pavement performance indicators for the corresponding years. The same roads were matched and combined to obtain a set of sequences of pavement performance changes with the features of the current year; with the recurrent-neural-network-based long short-term memory (LSTM) network and gate recurrent unit (GRU) network, the prediction accuracy of highway pavement performance on the test set was significantly increased. The prediction result indicates that the generalization ability of the improved recurrent neural network model is satisfactory, with the R2 achieving 0.936, and of the two models the GRU model is more efficient, with an accuracy thatreaches almost the same level as LSTMbut with the training convergence time reduced to 25 s. This study demonstrates that data generated by the work of maintenance units can be used effectivelyin the prediction of pavement performance. This article is part of the theme issue 'Artificial intelligence in failure analysis of transportation infrastructure and materials'.

Study on the Dynamic Performance of PU and SBS-Modified Asphalt Mixtures with Dense Gradation

A polyurethane (PU) mixture with excellent strength is regarded as an alternative to a modified asphalt mixture, but characteristic analysis on them is lacking. In this paper, the dynamic modulus of PU- and SBS-modified asphalt mixtures with the same gradation, aggregate type, and binder content was investigated and compared in terms of dynamic and viscoelastic properties. Compared with the SBS-modified asphalt mixture, the PU mixture with an extremely high dynamic modulus and reduced phase angle at high temperatures had lower temperature sensitivity, which allowed it to resist high-temperature deformation. While the phase angle did not show a statistically significant correlation, the dynamic modulus between the two mixtures did. The dynamic modulus and phase angle values of the PU mixture showed relatively small deviations and could be fitted to produce acceptable master curves, which exhibited obvious differences compared to those of the SBS-modified asphalt mixture. The PU mixture exhibited elastic properties during the test temperature range and, since its thermal rheological property is much smaller than that of the SBS-modified asphalt mixture, it is closer to viscoelastic material. This study provides an understanding of the PU mixture’s dynamic and viscoelastic properties, as well as material information for pavement design and performance prediction with PU mixture layers.

Predicting Rutting Development of Pavement with Flexible Overlay Using Artificial Neural Network

Pavement maintenance and repair is a crucial part of pavement management systems. Accurate and reliable pavement performance prediction is the prerequisite for making reasonable maintenance decisions and selecting suitable repair schemes. Rutting deformation, as one of the most common forms of asphalt pavement failures, is a key index for evaluating the pavement performance. To ensure the accuracy of the commonly used prediction models, the input parameters of the models need to be understood, and the coefficients of the models should be locally calibrated. This paper investigates the prediction of the rutting development of pavements with flexible overlays based on the data of the Canadian Long-Term Pavement Performance (C-LTPP) program. Pavement performance data that may be related to rutting were extracted from the survey of Dipstick for data analysis. Then, an artificial neural network (ANN) was adopted to analyze the factors affecting the rut depth, and to establish a model for the rutting development of pavements with flexible overlays. The results of the sensitivity analysis indicate that rutting is not only affected by traffic and climatic conditions, but it is also greatly affected by the thickness of the surface layer and voids in the mixture. Finally, a rutting evaluation index was provided to describe the rutting severity, and the threshold of the pavement maintenance time was proposed based on the prediction results. These results provide a basis for predicting rut development and pavement maintenance.

Prediction of asphalt pavement performance based on DEPSO-BP neural network

With the application of machine learning rapidly gaining popularity in computer science and other fields, neural network techniques have successfully simulated the performance of in-service pavements as they are efficient in predicting and solving nonlinear relationships and in dealing with uncertain large-area pavement problems. In this paper, we address the problem of the optimal timing of preventive maintenance of asphalt pavements to accurately predict the condition index (pavement condition index, PCI) of highway asphalt pavements and develop a highly accurate, long-period, multifactor prediction model with the suitability of preventive maintenance at its core. The prediction model is called differential evolution particle swarm optimization back propagation (DEPSO-BP) neural network, and the input dimension of the prediction model is determined by gray correlation analysis (GCA), and DEPSO is used to improve the search efficiency of BP neural network and the asphalt pavement usage performance with parameter continuity prediction model. Finally, the Qinglan Highway (G22) PCI of Gansu Province, China, is selected for example validation, and the prediction results are compared with those of the four models. The results show that the multifactor prediction model based on DEPSO-BP neural network has good generalization ability. This model is important for improving the economic efficiency of road maintenance, and can be used in the long-cycle process to provide model reference and scientific basis for the subsequent road maintenance budget application and decision-making scheme.