Utilizing spatial artificial intelligence to develop pavement performance indices: a case study.

Area of Science:

• Civil Engineering
• Computer Science
• Geographic Information Science

Background:

• Traditional pavement assessment methods are inefficient for modern transportation networks.
• Pavement Condition Index (PCI) and International Roughness Index (IRI) are key performance indicators.
• IRI data collection is more accessible than detailed pavement distress data.

Purpose of the Study:

• To develop an integrated machine learning (ML) and Geographical Information Systems (GIS) approach for pavement performance assessment.
• To classify International Roughness Index (IRI) to estimate Pavement Condition Index (PCI) for flexible pavements.
• To optimize pavement maintenance strategies through intelligent decision support.

Main Methods:

• Utilized machine learning algorithms including Artificial Neural Network (ANN), Adaptive Boosting (AdaBoost), Support Vector Machine (SVM), Decision Trees (DT), and Random Forest (RF).
• Collected 1042 International Roughness Index (IRI) data points using the TotalPave smartphone application.
• Applied feature selection, spatial visualization, and uncertainty analysis for robust evaluation.

Main Results:

• The Random Forest (RF) model achieved a superior accuracy of 99.9% and an F1-score of 99.
• Support Vector Machine (SVM) showed the lowest accuracy at 85.8% with an F1-score of 40.3.
• Identified critical pavement sections through spatial analysis and validated model reliability.

Conclusions:

• The integrated ML and GIS framework provides a scalable and intelligent decision-support tool for pavement management systems (PMS).
• This approach enables proactive maintenance planning and optimizes resource allocation for transportation infrastructure.
• The study demonstrates a transformative solution for evaluating and predicting pavement conditions effectively.