Machine learning (ML) is becoming a powerful tool for forecasting pavement performance and enhancing the efficiency of maintenance and resource planning. This study investigates the application of machine learning, using the random forest (RF) algorithm, to predict pavement condition index (PCI) over time. Manually assessed data from 2014 and 2018 were used to forecast the PCI for 2022, allowing the model to capture deterioration patterns across residential and non-residential streets in Skellefteå Municipality, Sweden. Model performance was evaluated using several performance metrics to assess flexibility and generalization across a diverse set of input variables, including pavement age (A), several distress mechanisms (D), previous PCI rating (S), and traffic data (T). For non-residential streets, models were further grouped by surface treatment type to evaluate RF’s predictive accuracy under varying maintenance conditions. Among residential street models, the RF model using A+D+S achieved the best results, with an R² of 0.78 and a correlation coefficient of 0.89 using the combined 2014 and 2018 dataset. For non-residential streets, incorporating traffic variables (T) further improved accuracy, with the RF(A+D+S+T) model reaching an R² of 0.79 and a correlation coefficient (ρ) of 0.88. While performance varied slightly across treatment types, the PCI rating assessed in 2018 (Status2018) and the weighted distress value (WDV) emerged as the strongest predictors. Furthermore, using the combined dataset (2014 and 2018) did not significantly improve prediction accuracy over using 2018 data alone. The findings demonstrate that RF models can effectively predict PCI and enhance decision-making in municipal pavement management.

This study investigates the influence of moisture and suction on the resilient modulus (M_R) of subgrade soils. The research employs suction-controlled Repeated Load Triaxial (RLT) tests on three sandy materials with varying fines content. The soil water retention curves (SWRC) for the three materials were obtained and allowed expediting suction equilibrium outside the triaxial chamber by controlling the water loss and thus inferring the yielded suction with the SWRC parameters. The results show that M_R increases with lower moisture content and higher suction. Two stress-based models and two moisture-based models are evaluated for predicting M_R. The findings indicate that stress-suction models provide a good fit for the silty sands, but only the model where suction is an independent variable suits all tested materials. Additionally, a proposed suction-based model demonstrates promising results. Overall, the study highlights the importance of considering both moisture and suction for accurate M_R characterisation of subgrade soils.

Pavement response monitoring (PRM) supports infrastructure management by evaluating pavement performance and optimizing maintenance strategies. This study experimentally validates a continuous PRM system using embedded geophones and accelerometers to measure pavement surface deflections under real traffic loads. The system is designed as a cost-effective and easy-to-install solution for continuous pavement deflection monitoring. Field experiments were conducted on the E16 Motorway in Borlänge, Sweden, a cold-climate region, where sensor signals were collected from passing heavy vehicles. Geophones and accelerometers measured pavement vertical vibration velocity and acceleration, respectively. Signal processing methods, including filtering, integration, and transformation, were applied to derive vertical pavement surface deflections. Measured deflections were compared with theoretical values calculated using the multilayer elastic response software ERAPave, showing low normalized root mean square error (NRMSE). Both sensor types accurately captured pavement surface deflections; however, geophones performed better, with an average NRMSE of 9.87% compared to 11.92% for accelerometers. Geophone measurements also showed close alignment between measured and calculated peak deflections, achieving an R² of 80%. This validation confirms the potential of embedded geophone and accelerometer setups as scalable, low-cost solutions for continuous PRM to support roadway maintenance and infrastructure durability.

This study applies sigmoidal models to predict pavement condition index (PCI) deterioration over time for residential and non-residential (main, collector and industrial) streets using pavement condition data from Skellefteå Municipality, Sweden. The dataset includes windshield-survey pavement assessments conducted in 2014, 2018 and 2022. Two modelling configurations were applied: temporal models calibrated using 2014 and 2018 observations to predict 2022 PCI ratings, and full dataset models developed using all available observations to represent overall deterioration behaviour. Separate sigmoidal curves were developed for residential and non-residential streets, as well as for the specific maintenance-treatment categories applied exclusively to non-residential streets. Sigmoidal curves for residential streets showed lower predictive performance and higher variability, potentially associated with heterogeneous pavement conditions, inconsistent pavement age and localized utility cuts, whereas sigmoidal curves for non-residential streets demonstrated comparatively better predictive capability due to more consistent deterioration patterns. Model performance also varied across maintenance-treated pavement surfaces, partly due to limited sample sizes. Sigmoid best-fit curves for street pavements treated with special treatments (ST) showed the highest predictive performance, while milling and resurfacing (MR)-treated pavements demonstrated relatively good agreement with measured PCI ratings. Surface levelling (SL) is a relatively low-cost treatment and may be suitable for earlier stages of deterioration, although traffic and pavement characteristics should also be considered when selecting treatments. Differences in unit cost between treatment types are considered only for contextual interpretation. Overall, sigmoidal curves may support more timely maintenance planning in data-limited municipal pavement management and reduce the need for major reconstruction.

Assuring sufficient structural stability of cycle paths is important to avoid distress. Being thin and narrow structures, they are highly susceptible to edge cracking by passings of heavy vehicles. Two important factors that affect distress are lack of lateral support at the edge and increased moisture content in the unbound granular materials. Three full-scale instrumented cycle path structures with different structural design were constructed in a concrete pit at a test facility. Response measurements were conducted with a falling weight deflectometer, a light weight deflectometer, and light, standard and heavy maintenance vehicles at three different groundwater table scenarios. The results confirm reduced load-bearing capacity close to the edge, and at an increased groundwater table. In general, larger transverse tensile strains can also be expected in the asphalt layer close to the edge.

Rutting, a prevalent failure mode in flexible pavements, largely stems from subgrade issues. Despite this, there is a lack of standard protocols to evaluate subgrade rutting or permanent deformation (PD). This study attempted to characterise PD in subgrades, focusing on a poorly graded sand and two silty sands. Moisture contents above and below optimum levels were considered to account for seasonal variations. The research involved adapting a test to assess the PD by determining typical stresses on the subgrade. Moreover, given these soils’ unsaturated state and medium- to fine-grained nature, suction is an important factor. Suction-controlled multi-stage repeated load triaxial tests were conducted, and the results were fitted by a PD model modified to account for suction. The characterisation was compared with the subgrade strain criterion used in pavement design solutions. Results indicated discrepancies between the PD characterisation and strain criteria predictions, with the silty sands performing better than the poorly graded sand, consistent with the shakedown theory.

Pavement condition prediction is essential for effective street maintenance. Reliable prediction models enable municipalities to anticipate pavement deterioration and plan maintenance proactively. This paper examines the use of machine learning (ML) models to enhance prediction accuracy for municipal streets. Seventy-two models were developed to predict the 2022 pavement condition index (PCI) using Linear Regression (LR), Random Forest (RF) and Neural Network (NN) algorithms. The study data was collected through windshield surveys in 2014, 2018 and 2022 in Skellefteå Municipality, Sweden. The models were trained and tested separately on the 2018 dataset and the combined 2014 and 2018 datasets, incorporating several combinations of variable sets for residential and non-residential (main, collector, and industrial) streets. Additional models were developed for non-residential streets categorized by maintenance treatments to analyse the predictive capability of the models. RF consistently outperformed other models, with the RF (A+D+S) model achieved the highest accuracy for residential streets, with a marginal improvement using the combined dataset. The RF (A+D+S+T) model performed best for non-residential streets and was also the most robust for pavement segments with surface levelling and milling & resurfacing. Among the variables, WDV (Weighted Distress), Status2018 (the PCI rating based on the 2018 assessment), LAR (longitudinal/transverse cracking), A (pavement age), LT (light traffic), and SU (surface unevenness) were the most significant contributors in the best-performing models. These findings provide municipalities with practical guidance for data-driven maintenance, helping to prioritize interventions and optimize resources. The insights also support informed decisions on maintenance frequency and treatment selection to extend pavement service life.

Machine learning (ML) models are increasingly getting attention in predicting pavement maintenance methods to improve decision-making. This study investigates the use of ML at the municipal level to predict the street pavement condition index (PCI) rating over a 4-year span. Several supervised learning models, namely linear regression (LR), random forest (RF), and neural network (NN), were applied to the visually assessed pavement condition data of Skellefteå municipality, Sweden. Pavement distress, pavement age, and traffic data were used in several combinations to evaluate and compare the performance of the models. The RF model was based on paired variables of pavement age and pavement distress data. The results were comparatively accurate with R2=0.59 and Spearman's coefficient=0.74 for residential streets in the model testing stage. Similarly, for main, collector, and industrial (MCI) streets, the RF model, based on pavement age and traffic variables, performed best with R2=0.79 and Spearman's coefficient=0.88 during the model testing stage. The importance of input variables varies with the level of the model's sophistication and pavement performance goal; however, pavement age is the dominant variable. The prediction models can be useful in effectively managing street networks among municipalities, even those with scarce resources.

One step towards reducing construction and industrial wastes (CIW) is to enhance their utilisation in road construction. Prior to their implementation, it is essential to evaluate their mechanical properties to ensure suitability and establish design guidelines. With this objective, four CIW materials were tested using repeated load triaxial (RLT) equipment to evaluate the materials’ functional properties, namely stiffness (Mr) and permanent deformation (PD) characteristics. In addition, full-scale accelerated pavement tests (APTs) were conducted where the four CIW materials were used as the subbase layer instead of traditional materials. Thereafter, a mechanistic–empirical (ME) pavement design and performance prediction method was used to predict their APT performances based on the RLT testing parameters. It was found that the ME performance method can adequately predict the CIW´s rutting development. The approach presented here, therefore, can facilitate the development of design guidelines and the increased utilisation of CIW-based materials in road construction projects.

Determining resilient modulus (MR) for subgrades is crucial in advancing Mechanistic-Empirical (M-E) pavement design. This study models the responses of a poorly graded sand and two silty sands through Suction-controlled Repeated Load Triaxial tests (RLT), factoring in moisture content and equivalent suction levels based on Soil Water Retention Curves (SWRC). Using a predictive model, the authors calculate the subgrade’s MR for a typical pavement cross-section in Sweden, incorporating layer parameters for critical seasons and climatic zones defined by the Swedish Transport Administration (STA). Results highlight the materials' sensitivity to moisture-suction and their stress dependency. Notably, the predicted MR for silty sands exceeded STA-recommended values across seasons and climatic zones. The poorly graded sand aligns well when the coefficient of earth pressure at rest k0 equals 1, except under wet conditions, in such case STA-recommended values are optimistic. Comparisons with existing data support the findings, particularly for the silty sands. In summary, this research sheds light on three subgrades and offers a reproducible method to expand the database of subgrade materials. Furthermore, if offer insights for enhancing M-E pavement design, considering different climatic conditions and materials.