Long-Term Pavement Performance Database Research Articles

Development of a Machine Learning (ML)-Based Computational Model to Estimate the Engineering Properties of Portland Cement Concrete (PCC)

AbstractPortland cement concrete (PCC) is the construction material most used worldwide. Hence, its proper characterization is fundamental for the daily-basis engineering practice. Nonetheless, the experimental measurements of the PCC’s engineering properties (i.e., Poisson’s Ratio -v-, Elastic Modulus -E-, Compressive Strength -ComS-, and Tensile Strength -TenS-) consume considerable amounts of time and financial resources. Therefore, the development of high-precision indirect methods is fundamental. Accordingly, this research proposes a computational model based on deep neural networks (DNNs) to simultaneously predict the v, E, ComS, and TenS. For this purpose, the Long-Term Pavement Performance database was employed as the data source. In this regard, the mix design parameters of the PCC are adopted as input variables. The performance of the DNN model was evaluated with 1:1 lines, goodness-of-fit parameters, Shapley additive explanations assessments, and running time analysis. The results demonstrated that the proposed DNN model exhibited an exactitude higher than 99.8%, with forecasting errors close to zero (0). Consequently, the machine learning-based computational model designed in this investigation is a helpful tool for estimating the PCC’s engineering properties when laboratory tests are not attainable. Thus, the main novelty of this study is creating a robust model to determine the v, E, ComS, and TenS by solely considering the mix design parameters. Likewise, the central contribution to the state-of-the-art achieved by the present research effort is the public launch of the developed computational tool through an open-access GitHub repository, which can be utilized by engineers, designers, agencies, and other stakeholders.

Ensemble Learning Approach for Developing Performance Models of Flexible Pavement

This research utilizes the Long-Term Pavement Performance database, focusing on devel-oping a predictive model for flexible pavement performance in the Southern United States. Analyzing 367 pavement sections, this study investigates crucial factors influencing asphaltic concrete (AC) pavement deterioration, such as structural and material components, air voids, compaction density, temperature at laydown, traffic load, precipitation, and freeze–thaw cycles. The objective of this study is to develop a predictive machine learning model for AC pavement wheel path cracking (WpCrAr) and the age at which cracking initiates (WpCrAr) as performance indicators. This study thoroughly investigated three ensemble machine learning models, including random forest, extremely randomized trees (ETR), and extreme gradient boosting (XGBoost). It was observed that XGBoost, optimized using Bayesian methods, emerged as the most effective among the evaluated models, demonstrating good predictive accuracy, with an R2 of 0.79 for WpCrAr and 0.92 for AgeCrack and mean absolute errors of 1.07 and 0.74, respectively. The most important features influencing crack initiation and progression were identified, including equivalent single axle load (ESAL), pavement age, number of layers, precipitation, and freeze–thaw cycles. This paper also showed the impact of pavement material combinations for base and subgrade layers on the delay of crack initiation.

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.

Modelling the effects of flexible pavement distresses in the long-term pavement performance database on performance

Evaluating flexible pavement performance is mandatory for managing transport infrastructure. This study focuses on modeling the relationships between international roughness index (IRI) and a total of 10 types of pavement distress, including alligator, block, wheel path length, wheel path longitudinal, non-wheel path longitudinal, transverse crackings, patches, bleeding, raveling areas, and pumping. The data recorded under the Long-Term Pavement Performance was used to develop the models. Data sets covering General Pavement Studies from seven states of the United States were used in modeling. The study used modeling approaches, including nonlinear regression analysis, multivariate adaptive regression splines, and artificial neural networks (ANN), in which IRI was the dependent variable and pavement distress was the independent variable. In the developed models, 0.516, 0.623, and 0.684 regression coefficients values were obtained for nonlinear regression analysis, multivariate adaptive regression splines, and artificial neural networks approaches, respectively. The analysis results have determined that the artificial neural networks technique performs more successfully than the other techniques. The statistical error analyses of the root mean square error, Nash-Sutcliffe coefficient of efficiency, mean absolute error, and normalized root mean square error also showed that the same modeling approach performs more successfully. With these data generated from a universally used database, it has been determined once again that ANN is the most efficient mathematical approach in modeling the relationships between surface distresses and IRI.

Performance-oriented road structure and material design method based on enhanced XGBoost algorithm

ABSTRACT A well-designed road structure is critical for reducing road operational period diseases and saving construction management and maintenance funds throughout the life cycle of roads. Excessive design may cause waste in materials or pavement distresses. This study proposes an intelligent optimisation design method for road structure and materials based on pavement performance and cost. The Long-Term Pavement Performance (LTPP) database were used to training the prediction model using the particle swarm optimisation (PSO) algorithm and the extreme gradient boosting algorithm (XGBoost) algorithm. Subsequently, an automated design model encompassing pavement thickness, pavement material, and subgrade material was established based on the predictive model derived from the PSO algorithms. The results indicate that the R2 between the actual values and the predicted values of the PSO-XGBoost model is above 0.935, and the Pareto optimal solution is obtained using the multi-objective Particle Swarm Optimisation algorithm. This investigation showcases the potential of data-driven methodologies in furnishing valuable guidance for the initial stages of road design and construction.

Optimizing Faulting Prediction for Rigid Pavements Using a Hybrid SHAP-TPE-CatBoost Model

Faulting refers to the common and significant distress in Jointed Plain Concrete Pavement (JPCP), which has an adverse impact on the pavement roughness. Nevertheless, the existing fault prediction models continue to heavily rely on conventional linear regression techniques or basic machine learning approaches, which leaves room for improvement in training efficiency and interpretability. To enhance training efficiency and accuracy, this study developed five novel faulting prediction models. These models are based on five basic machine learning algorithms: Random Forest (RF), Additive Boosting (AdaBoost), Gradient Boosting Decision Tree (GBDT), Light Gradient Boosting Machine (LightGBM), and Categorical Boost (CatBoost), combined with the tree-structured Parzen estimator (TPE). The five models are TPE-RF, TPE-AdaBoost, TPE-GBDT, TPE-LightGBM, and TPE-CatBoost. In addition to selecting the best-performing model, this study incorporated the Shapley Additive Explanation (SHAP) technique and developed TPE-SHAP-CatBoost to improve the interpretability of the model’s predictions. The process involved extracting historical data on pavement performance, including 17 variables, from the Long-Term Pavement Performance (LTPP) database for 160 instances of observation. Firstly, the Boruta method was used to identify the final set of input variables. Secondly, the TPE technique, which is a Bayesian optimization method, was applied to automatically select the optimal hyperparameters for the base models. Finally, SHAP was used to provide both global and local explanations of the model’s outputs. The results indicate that the TPE-CatBoost model achieves the highest accuracy with an R2 value of 0.906. Furthermore, the TPE-SHAP-CatBoost model identified the primary factors influencing faulting by incorporating SHAP and provided explanations of the model’s results at both the global and local levels. These research findings highlight the ability of the proposed model to accurately predict faulting, providing precise and interpretable guidance for pavement maintenance while reducing workload for pavement engineers in data collection and management.

A reliability-base method for thermal cracking prediction in asphalt concrete

This study focuses on understanding the impact of environmental factors that contribute to the deterioration and damage of Hot Mix Asphalt (HMA) pavements. One significant problem faced by HMA pavements is transverse (thermal) cracking, which refers to the formation of cracks perpendicular to the centerline of the pavement. Over the past few decades, researchers have put significant effort into analyzing and understanding how thermal cracking affects HMA pavements. Mechanistic-empirical models play a crucial role in this process, but they require a substantial amount of input data related to the pavement's structural and material properties. Obtaining this data can be done through various methods with different levels of quality, which affect the reliability of the pavement performance responses. The probabilistic base designs and specifications can consider the variation of input data and the impact of material and structural together with environmental parameters on pavement deterioration. In this project, transverse cracking, as one of the major asphalt concrete pavement cracking, particularly in the freezing areas, will be studied, and a reliability-based method will be introduced. To reach this goal, the long-term pavement performance (LTPP) database was used. A regression model from the key parameters of pavement material and structural properties and environmental and traffic indicators was generated. Then through the structural reliability method, the probability of overpassing of transverse cracking amount from the threshold for a pavement design was assessed.

Developing an improved extreme gradient boosting model for predicting the international roughness index of rigid pavement

The International Roughness Index (IRI) is a crucial indicator for evaluating pavement performance. However, existing prediction models for IRI in rigid pavements primarily rely on linear regression or simple machine learning techniques, which necessitates improvements in both training efficiency and interpretability. In addressing this issue, a proposed IRI prediction model combines the Extreme Gradient Boosting (XGBoost) method with advanced hyperparameter optimization utilizing the Tree-structured Parzen Estimator (TPE) technique. Furthermore, the model incorporates the Shapley Additive Explanations (SHAP) interpretability framework. To train and validate the model, 146 historical observation records were extracted from the Long-Term Pavement Performance database. The Boruta algorithm was employed to select the final input variables. Results demonstrated that after TPE hyperparameter optimization, the XGBoost model achieved an R2 value of 0.896, and an RMSE value of 0.187. Surpassing performance compared to other ensemble learning models. The Boruta algorithm successfully identified all significant variables, thereby improving the model's accuracy on the testing set. Additionally, the SHAP framework confirmed the primary factors influencing IRI and provided explanations for the model's results at both global and local levels. These findings highlight the ability of the proposed model to accurately predict IRI, providing precise guidance for pavement maintenance while reducing the workload associated with data collection and management for pavement engineers.

Implementation of deep neural networks and statistical methods to predict the resilient modulus of soils

ABSTRACT The Resilient Modulus (Mr) is perhaps the most relevant and widely used parameter to characterise the soil behaviour under repetitive loading for pavement applications. Accordingly, it is a crucial parameter controlling the mechanistic-empirical pavement design. Nonetheless, determining the Mr by laboratory tests is not always possible due to the high consumption of time and financial resources. Thus, developing new indirect approaches for estimating the MR is necessary. Precisely, this article investigates the application of Deep Neural Networks (DNNs) and statistical methods to predict the Mr of soils. For that purpose, the Long-Term Pavement Performance (LTPP) database was implemented. It includes 64 701 datasets resulting from coarse-grained and fine-grained soil samples considering a wide range of grain size distribution and subjected to different stress levels. The input parameters were the bulk stress, octahedral shear stress, and the percentage of soil particles passing through the different sieves (3”, 2”, 3/2”, 1”, 3/4”, 1/2”, 3/8”, No. 4, No. 10, No. 40, No. 80, and No. 200) and the output was the Mr. The results suggest that while conventional mathematical models are unable to predict the influence of the grain size distribution and stress level on the Mr, the proposed DNNs were able to reproduce very accurate predictions. Notably, the proposed computational models have been uploaded to a GitHub repository and have become a valuable tool for forecasting the Mr when experimental measurements are not feasible.

Creating Rutting Prediction Models through Machine Learning Techniques Utilizing the Long-Term Pavement Performance Database

Over time, roads undergo deterioration caused by various factors such as traffic loads, climate conditions, and material properties. Considering the substantial global investments in road construction, it is crucial to periodically assess and implement maintenance and rehabilitation (M and R) plans to ensure the network’s acceptable level of service. An integral component of the M and R plan involves utilizing performance prediction models, especially for rutting distress, a significant issue in asphalt pavement. This study aimed to develop rutting prediction models using data from the Long-Term Pavement Performance (LTPP) database, employing several machine learning techniques such as regression tree (RT), support vector machine (SVM), ensembles, Gaussian process regression (GPR), and Artificial Neural Network (ANN). These techniques are well-known for effectively handling extensive and complex datasets. To achieve the highest modeling accuracy, the parameters of each model were meticulously fine-tuned. Upon evaluation, the results revealed that the GPR models outperformed other techniques in various metrics, including Root Mean Square Error (RMSE), R-squared, Mean Absolute Error (MAE), and Mean Square Error (MSE). The best GPR model achieved an RMSE of 1.96, R-squared of 0.70, MAE of 1.32, and MSE of 109.33, indicating its superior predictive capabilities compared with the other machine learning methods tested in this study. Comparison Analysis was made for 10 randomly selected sections on our novel machine learning model that outperforms existing models, with an R2 of 0.989 compared with 0.303 and 0.3095 for other models. This demonstrates the potential of advanced machine learning in accurate rut depth prediction across diverse climates, aiding pavement management decisions.

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'.

Computationally-Affordable Unsupervised Machine Learning Algorithm to Identify the Level of Distress Severity in Pavement Functional Performance

Pavements continually experience various types of distresses that are manifested in the form of defects that will eventually lead to the pavement’s failure. Selection of appropriate preventive maintenance strategies in terms of timing and method retards the large backlog of poor conditions of pavements, therefore improves the system-wide performance in an efficient and cost-beneficial way. However, the pre-treatment functional performance condition of asphalt pavement has a significant impact on the efficacy of preventive maintenance, and to provide years of service, it is necessary to designate an appropriate preventive maintenance method according to the actual functional condition of pavement. This study presents a new unsupervised learning approach for 1) extracting high correlation factors to reduce the dimensionality of feature distress data, and 2) accurately separating attributes that define pavement functional performance into severity levels to effectively assess the practiced preventive maintenance. Three unsupervised machine learning algorithms including feature extract, relevant factor analysis, and cluster analysis were executed. Based on the severity level results of the pre-treated pavement, the effects of four treatment methods (thin overlay, slurry seal, chip seal, and crack seal) were evaluated by a statistical analysis. The preventive maintenance data from the Specific Pavement Studies-3 (SPS-3) experiment of the Long-Term Pavement Performance (LTPP) database are used for the analysis. The results demonstrated that selecting the appropriate maintenance method for the functional performance risk level of the pavement is beneficial in achieving successful treatment results.

Field Aging Characterization of Asphalt Pavement Based on the Artificial Neural Networks and Gray Relational Analysis

Accurate characterization of field aging of asphalt pavement is critical to precisely assessing its in-service performance. However, most of the traditional test/predictive methods either cannot fully capture the field aging characteristics or involve costly testing/computational efforts to ensure satisfactory prediction accuracy. To alleviate these problems, this study developed a new field aging predictive model based on artificial neural networks (ANNs) and gray relational analysis (GRA), which takes the field-aged viscosity of asphalt binder as the target predictive property. A series of influencing factors that may affect the field-aged viscosity were systematically investigated, among which the eight most significant ones were screened out for the ANN modeling through the GRA. A total of 479 data extracted from long-term pavement performance (LTPP) database were used for the training, validation, and testing of the ANN model. The calculation results showed that the predictive model developed using the ANN approach provided a high prediction accuracy with R2 value greater than 0.90. Furthermore, the falling-weight deflectometer (FWD) data collected from the database were utilized to evaluate the predictive performance of the well-trained ANN model. Consistent results were obtained between the viscosity values predicted from the ANN model and those back-calculated from the FWD data, indicating that the newly developed field aging model has the capability to accurately characterize the field aging evolution of asphalt pavement.

International Roughness Index Prediction for Jointed Plain Concrete Pavements Using Regression and Machine Learning Techniques

Rigid pavements are recently gaining vast interest for their durability and large traffic volume endurance. This raises the need for a robust method to predict their long-term performance to support decisions on their future maintenance and rehabilitation requirements. One of the most descriptive parameters of pavement ride quality is the International Roughness Index (IRI). IRI can be used as a performance indicator for the level of serviceability of rigid pavement. This study presents the evaluation of 21 variables covering climate, traffic, and structural parameters in predicting IRI for jointed plain concrete pavement (JPCP) sections using 1,414 data points acquired from the Long-Term Pavement Performance (LTPP) database. It was found that only 10 variables are significant in the IRI prediction for JPCP. Moreover, four modeling techniques were applied, namely, linear regression, multivariate adaptive regression splines, Gaussian process regression, and artificial neural network. It was found that a deep ANN model with a structure of one input layer, three hidden layers, and one output layer [10-36-18-9-1] gives the highest IRI prediction accuracy with root mean square error and coefficient of determination (R2) of 0.117 and 0.92, respectively. Finally, a sensitivity analysis was performed to evaluate the impact of each input variable on the accuracy of IRI prediction for JPCP sections, and it was found that the most influential variables are initial IRI, pavement age, mean and standard deviation of humidity, standard deviation of evaporation, mean and standard deviation of freezing index, pavement compressive strength, ratio between pavement and base layers thickness, and finally an estimated parameter called site factor which is a function of (age, freezing index and soil percentage passing sieve no. 200).

Evaluating the impact of overweight trucks on rigid pavement performance using AASHTOWare pavement ME design

ABSTRACT The purpose of this study is to evaluate the impact of overweight vehicles on pavement performance over its design life, with Jointed Plain Concrete Pavements (JPCPs) particularly considered for analysis. The Long-Term Pavement Performance database was used to provide information on pavement construction, age, overall structure, specific traffic inputs, and recorded historical pavement performance. AASHTOWare Pavement ME Design (PMED) was used to analyze data extracted from LTPP based on legislation by the Federal Highway Administration (FHWA) regulating gross vehicle weight (GVW) and single and tandem axle weights of overweight trucks (OWTs) to estimate pavement performance and service life. The overall findings indicated that although increasing overweight by 20% on single or tandem axles could considerably reduce pavement service life by up to 20% in the various scenarios examined in this study, OWT exceeding the GVW limit legally-allowed by FHWA could damage pavement more than OWT exceeding axle weight limits.

Impact of sea level rise on asphalt pavement responses considering seasonal groundwater and moisture gradient in subgrade

Sea level rise (SLR) has been found to have a great impact on the transportation infrastructure in coastal areas. This study developed an analysis approach with numerical modeling to analyze the impact of SLR-induced groundwater rise on critical responses of coastal pavements. The numerical model employed a stress-, moisture content- and matric suction-dependent modulus model to consider the unsaturated characteristics of subgrade soil and non-linear distribution of moisture content. The model results were validated using Falling Weight Deflectometer (FWD) tests in the field. Based on the realistic moisture content collected from Long-Term Pavement Performance (LTPP) database, the critical responses of pavements were evaluated. The in-situ moisture content is non-linearly distributed along the depth, which differs from traditional assumption of monotonic distribution. The seasonal groundwater table (GWT) and subgrade soil type are two factors impacting the critical pavements responses under SLR scenarios. The pavement will be significantly impacted by SLR in the seasons with the higher GWT. The modulus of clay soil is more sensitive to moisture content than that of sandy or silty soil and thus causes the greater increase of pavement responses as sea level rises. On the other hand, the compressive stain on subgrade increases more remarkably than tensile strains at the bottom of asphalt layer under the impact of SLR.

CALIBRATION OF REGRESSION-BASED MODELS FOR PREDICTION OF TEMPERATURE PROFILE OF ASPHALT LAYERS USING LTPP DATA

For analysis, design, and rehabilitation purposes of flexible pavements, the temperature profile of asphalt layers should be determined. The predictive models as an alternative to in-situ measurements, are rapid and easy methods to determine the temperature of asphalt layer at various depths. These models are developed based on limited field data. Hence, there is a need for developing new models for prediction of temperature profile of asphalt layers in various climatic regions. In this study, climatic data was retrieved from the Long-Term Pavement Performance (LTPP) database. The information of 33 asphalt pavement test sections in 16 states in the United States was employed for calibrating the predictive models. Using the prepared data, the temperature profile of asphalt layers was predicted utilizing four regression-based models, including Ramadhan and Wahhab, Hassan et al., Albayati and Alani, and Park et al. models. Existing prediction models were calibrated, and to predict the temperature profile of asphalt layer, new models were developed. Performance evaluation and validation of newly developed models showed an excellent correlation between predicted and measured values. Results show the ability of the developed models in predicting the temperature profile of asphalt layers with very good prediction precision (R2 = 0.94) and low bias.

Application of Artificial neural network technique for prediction of pavement roughness as a performance indicator

One of the most important and widely accepted pavement performance and ride quality indicators is the International Roughness Index (IRI). This study investigates the combined effect of pavement distress on flexible pavement performance in two climate regions (wet freeze and wet freeze) in the U.S. and Canada. The long-term pavement performance (LTPP) database was used to obtain pavement distress data. Data from forty-three of the LTPP pavement sections (333 observations) with no previous maintenance were collected. The proposed models predict the IRI as a function of pavement distress variables, namely the pavement age, rutting, fatigue cracking, block cracking, longitudinal cracking, transverse cracking, potholes, patching, bleeding, and ravelling. After the data were collected, modelling was conducted to predict IRI using two techniques: multiple linear regression (MLR) and artificial neural network (ANN). The coefficient of determination (R2), root mean squared error (RMSE) and mean absolute error (MAE) were used to examine the performance of the two techniques adopted in this study. The models' results revealed that both ANN and MLR models could predict IRI with good accuracy. The MLR models yielded the R2 values of 77.7% and 89.3%, whereas the ANN models resulted in the R2 values of 99.1% and 97.5% for wet freeze and wet no freeze climate regions, respectively. As a result, ANN models are more accurate and efficient than MLR models.

Development of an Artificial Neural Network-Based k-Value Prediction Model to Improve the Sensitivity of Base Layer on Rigid Pavement Performance

The performance of rigid pavement is greatly affected by the properties of the base/subbase and subgrade layer. However, the performance predicted by the AASHTOWare Pavement Mechanistic-Empirical (ME) Design shows low sensitivity to the properties of base and subgrade layers. In this study, a new set of modified models, that is, resilient modulus ( MR) and modulus of subgrade reaction ( k-value) are adopted to better reflect these layers’ influence on the overall performance. An artificial neural network (ANN) model is developed to predict the modified k-value based on finite element (FE) analysis. The training and validation data sets in the ANN model consist of 27,000 simulation cases with different combinations of pavement layer thickness, layer modulus, and slab-base interface bond ratio. Eight pavement section data are collected from the long-term pavement performance (LTPP) database and modeled using the FE software ISLAB2000 to evaluate the sensitivity of the modified MR model and k-values to pavement performance. The computational results indicate that the modified MR values have a higher sensitivity to water content in the base layer on fatigue cracking (bottom-up and top-down cracking) and the faulting performance of rigid pavements compared with the results using the Pavement ME design model. It is also observed that the k-values using the ANN model have a greater sensitivity to cracking and faulting performance as a result of changes in any bonded conditions (fully bonded, partially bonded, zero bonded), whereas the Pavement ME design model can only calculate at two extreme bonding conditions (i.e., fully bonded and zero bonded).

Analysis on Effects of Joint Spacing on the Performance of Jointed Plain Concrete Pavements Based on Long-Term Pavement Performance Database.

With the day-night temperature and moisture levels changing every day, expansion and shrinkage of concrete slabs is always occurring; therefore, joints provide extra room for concrete slab deformation. The joint spacing in jointed plain concrete pavement (JPCP) is continuously affecting long-term pavement behaviors. In this study, data from the Long-Term Pavement Performance (LTPP) program were analyzed, and the behaviors of JPCP with different joint spacings were compared to discover the joint spacing effects. Since LTPP has an enormous database, three representative sections located in different states were selected for analysis, where the variable factors such as temperature, moisture, and average annual daily truck traffic (AADTT) were almost the same between the three sections. Three different joint spacings, including 15 ft (4.5 m), 20 ft (6 m), and 25 ft (7.5 m), were compared based on the collected LTPP data. The involved long-term pavement performances, such as average transverse cracking (count), average JPCP faulting, international roughness index (IRI), and falling weight deflectometer (FWD) deflections were compared between JPCP with different joint spacings. Based on the comparative analysis, the JPCP constructed with a 15 ft joint spacing demonstrated the best long-term performance. It showed no transverse cracking, the lowest average JPCP faulting, the best IRI value, and the smallest FWD deflection during the entire in-service period. With proper joint spacing, the cost of road maintenance throughout the life cycle could be significantly reduced due to there being less distress. Therefore, it is recommended to optimize the joint spacing to about 15 ft in JPCP in future applications.