Term 252

AI-Driven Road Inspection and Quality Reporting System

Project Type: Self-Initiated

Project Description

This project proposes an AI-driven road inspection and quality reporting system to accelerate the detection of roadway incidents and reduce inspection time, cost, and the risk of defects worsening over time. The system targets three primary inspection tasks: (1) detecting potholes and pavement defects, (2) identifying damaged or missing road signs, and (3) assessing poorly painted or degraded road markings. The proposed concept is a mobile robotic platform equipped with vision sensors (camera) and LiDAR to support robust perception and localization. At the conceptual design phase, the system will integrate computer vision and deep learning for defect detection, alongside autonomous navigation and localization supported by reliable communications and signal-processing techniques for efficient data transfer. Inspection outputs will be processed and transmitted to a central data center, where results are presented through a web-based monitoring dashboard backed by an SQL database. The dashboard will display near–real-time inspection results and robot status, including defect type, severity indicators, geolocation, and supporting sensor evidence. A solar-powered docking/warehouse station is also included to enable autonomous recharging and continuous operation.