Abstract

The Surface Distress Index (SDI) is a key parameter for assessing urban road conditions that is effective in sustainable infrastructure management. The current research gap focuses on high-quality roads and the absence of predictive models applicable to lower-quality infrastructure, while complex maintenance is often overlooked, especially on urban roads with diverse types of surface damage. The objective of this study is to develop a predictive model of the Surface Distress Index (SDI) based on Artificial Neural Networks (ANN) to enhance road maintenance planning in urban areas. This model was trained using five years of urban road damage data from 42 city road segments. The coefficient of determination from the research results indicates a very high prediction accuracy, with an R value of 0.99, the MAE of 0.01, and the RMSE of 0.03. This model offers a more dynamic plan to enhance the sustainable maintenance of urban infrastructure. The resulting predictive model provides adaptive solutions to existing problems, environmental changes, and supports more sustainable urban infrastructure management.