The condition of U. S. road infrastructure continues to deteriorate, with over 39% of public roadways rated in poor or mediocre condition and an annual economic impact exceeding 160 billion. Traditional visual inspection methods are slow, laborintensive, and often subjective, while high-end survey vehicles and smartphone-based alternatives either lack scalability or measurement reliability. To address this gap, we present a scalable, AI-assisted approach for automated pavement condition monitoring using visual data collected from city-wide deployments. The proposed framework detects surface defects and assigns standardized condition scores aligned with industryaccepted rating systems. Field testing in two Indiana cities—West Lafayette and Fort Wayne—spanned over 2, 000 miles and resulted in more than 13 million geo-referenced image captures. Results demonstrate the system’s potential to support timely, consistent, and cost-effective evaluations of roadway health. This work highlights the importance of integrating AI with data-driven inspection practices to improve the sustainability and efficiency of municipal pavement management.
Sankar et al. (Tue,) studied this question.