Classification of Urban Road Damage Levels Using Surface Distress Index Parameters and K-means Algorithm

Abstract

Road damage is a significant problem in many countries, including Indonesia, and it can affect safety, transportation efficiency, and the quality of life of communities, especially on urban roads. This study aims to develop a data-based model using the K-Means algorithm to detect and classify the levels of urban road damage based on the Surface Distress Index (SDI) parameter. The data used comprised 2,467 road segments containing information on the types and levels of damage over the past five years. The clustering model was designed with two and four clusters, and the results indicated that the four-cluster model provided a clearer and more representative separation of road conditions. The Silhouette Coefficient value of the four-cluster model is 0.513, indicating a more detailed and clearer separation compared to the two-cluster model with a Silhouette value of 0.423. The four cluster model is better at telling apart complex data structures than the two cluster model. The results of this study contribute to the development of a road condition monitoring system based on maintenance priority data categorized by road condition, which can be adapted to improve infrastructure policies in Indonesia and other developing countries.