Pavement and Road Health Monitoring Using Random Forest Technique

Abstract

Detecting road anomalies is crucial for preventing road accidents, yet the challenges in achieving efficient detection persist. Traditional visual inspections are costly, cumbersome, and often unreliable, making them impractical for comprehensive road network monitoring. To overcome these issues, this study leverages machine learning (ML) techniques to develop an algorithm for road anomaly detection using acceleration data. The proposed algorithm employs the random forest (RF) technique for anomaly detection and classification, and its performance is compared with other models, including decision tree (DT), K-nearest neighbors (KNN), and support vector machine (SVM). The results highlight the superiority of the random forest (RF) classifier, achieving an impressive accuracy rate of 90.6%. Additionally, the SVM model emerges as a significant contender with an accuracy rate of 89.43%. This research underscores the potential of ML in enhancing road safety by efficiently and accurately identifying road anomalies, thereby reducing accident risks. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2026.