Smart Road Infrastructure: Pothole Detection with Real-Time Map Integration for Reconstruction Planning

Abstract

Maintaining Quality road infrastructure is necessary in promoting road safety and minimizing damages and repair costs incurred. Potholes are among the major problems that affect road infrastructures due to the effects of wear and tear. Potholes are not only costly to address but cause travel delays, accidents, and increased expenses. Traditional road inspection procedures have been inefficient and expensive for municipalities, making it impossible to monitor potholes for repair purposes effectively. In light of rapid urbanization, there has been a high demand for an efficient, inexpensive, and accurate method for road inspection and maintenance management. Therefore, this paper highlights an advanced pothole detection system featuring mapping and GPS functionalities. Deep learning-based pothole detection models, including YOLOv11, help detect and count potholes in a video file collected on various roads. The system estimates the total length of roads and marks potholes on the visual map. Users can easily monitor pothole distribution, road length and access reports in CSV and PDF formats from a dashboard.