Automatic inspection of cement pavement health condition using GPR

About the Project

The health and structural integrity of cement pavements are critical to ensuring road safety, longevity, and cost-effective maintenance. Traditionally, the inspection of a pavement’s condition relies on labor-intensive manual methods, which can be time-consuming, costly, and prone to human error. To address these challenges, this project proposes the use of a dual-polarized Ground Penetrating Radar (GPR) system combined with Artificial Intelligence (AI) techniques for the automatic, non-destructive inspection of cement pavements. This innovative approach will enable the detection and evaluation of key pavement health indicators, including inner cracks, rebar corrosion, layer debonding, and voids (caves), thereby optimizing maintenance decision-making processes and extending pavement lifespan.

The primary goal of this project is to develop an automated methodology for the comprehensive inspection of cement pavements, capable of identifying and assessing critical structural issues such as cracks, corrosion, debonding, and voids.

Methodology

1. Dual-Polarized GPR Technology: The dual-polarized GPR system utilizes two different electromagnetic wave polarizations (HH and VV) to scan the pavement at varying angles. By transmitting and receiving radar signals in both polarizations, the system can better identify a range of defects, including those that may be difficult to detect with conventional GPR systems.
   Inner Cracks: The system can distinguish cracks based on their size, depth, and orientation relative to the pavement surface.
   Rebar Corrosion: Corrosion of steel reinforcements within the cement can be identified by analyzing the GPR reflections and polarimetic analysis.
   Layer Debonding: GPR data can be used to detect delamination or debonding between concrete layers, which is a common issue in multi-layer pavements.
   Cave Formation (Voids): Voids or cavities within the pavement structure, which can lead to premature pavement failure, are identified by anomalous radar reflections from air gaps.
2. Artificial Intelligence (AI) for Data Processing and Analysis: The radar data captured by the dual-polarized GPR system will be processed using AI-based algorithms – particularly machine learning (ML) and deep learning (DL) models. The AI system will be trained on large datasets consisting of labeled GPR scans, where different types of pavement defects have been manually identified. Key tasks of the AI component include:
   Defect Detection and Classification: The AI model will automatically detect and classify defects based on the radar signal characteristics. Using convolutional neural networks (CNNs) or other suitable architectures, the system will classify defects such as cracks, corrosion, and voids.
   Defect Localization and Quantification: The AI training detector will also localize defects within the pavement and estimate their size, depth, and severity, providing detailed information about the condition of the road.
   Condition Assessment: By analyzing patterns in the data, the AI system will assess the overall health of the pavement, including predicting areas that are likely to fail in the near future based on detected defects and their severity.
3. System Integration and Automation: The dual-polarized GPR system will acquire a large amount of dual-polarized reflection data (HH and VV). The AI-driven data analysis will be performed in real-time or post-processing, enabling immediate feedback. The inspection system will be capable of running autonomously, with minimal human input, thereby increasing inspection efficiency, reducing costs, and enabling the frequent monitoring of pavement conditions.

Benefits:

1. Non-Destructive and Comprehensive Inspection: The dual-polarized GPR system provides a more accurate and detailed understanding of the pavement's internal structure by way of a non-destructive approach, allowing frequent and detailed inspections without affecting road traffic or requiring road closures.
2. Early Detection of Structural Issues: Early detection of defects such as cracks, corrosion, debonding, and voids can prevent more severe damage and costly repairs in the future. The AI system’s ability to assess the severity of these issues helps prioritize repairs based on urgency and severity, ensuring the more efficient application of maintenance resources.
3. Cost and Time Efficiency: Automating the inspection process significantly reduces labor costs and time compared to manual surveys. The system's ability to rapidly scan and analyze large sections of pavement allows maintenance agencies to inspect more roadways in less time, improving overall infrastructure management.
4. Enhanced Data Accuracy and Objectivity: The AI-driven analysis eliminates human subjectivity, ensuring that all detected defects are accurately identified and classified based on objective data. This increases the reliability of pavement condition assessments and helps prioritize repair efforts.