Multi-Method Non-Destructive Testing for Improving Bridge Health using AI for Proactive Structural Health and Predictive Maintenance

Abstract

Aging bridge infrastructure poses a growing challenge to public safety, resource management, and structural integrity, highlighting the urgent need for effective, non-invasive monitoring solutions. Traditional inspection methods often lack the accuracy, efficiency, and real-time capabilities required for proactive maintenance. This study examines four non-destructive testing (NDT) techniques—Ultrasonic Testing (UT), Ground-Penetrating Radar (GPR), Infrared Thermography (IRT), and Acoustic Emissions (AET)—to evaluate their respective strengths, limitations, and suitability for detecting various types of bridge deterioration. By testing each method on multiple bridge structures, we assess accuracy in detecting cracks, voids, and other common issues. Findings indicate that UT is highly effective for internal flaw detection, GPR for subsurface conditions, IRT for surface degradation, and AET for realtime crack monitoring. To overcome the limitations of single method monitoring, this study further explores a multi-method NDT system that combines all four techniques. Our integrated model significantly improves detection accuracy by leveraging the unique strengths of each method, enabling a more comprehensive assessment of bridge health. Additionally, artificial intelligence (AI) enhances this system’s predictive capabilities, offering real-time analysis and enabling predictive maintenance. Through AI-driven data fusion, infrastructure managers can shift from reactive to proactive strategies, thereby reducing maintenance costs, improving resource allocation, and extending bridge lifespan. Field trials demonstrate the integrated system’s potential to provide early-stage issue detection, enhance structural resilience, and promote long-term infrastructure sustainability. This combined approach provides a forward-looking solution for bridge management, supporting public safety and sustainable maintenance practices.