Detection of Hidden Defects in Urban Roads Using Ground-Penetrating Radar: Application and Case Study

Accurate identification and risk assessment of hidden defects in urban roads are essential for preventing road collapse and ensuring infrastructure safety. Based on a large-scale investigation of urban roads in Hunan Province, China, this study proposed a multi-scale collaborative detection framework integrating three-dimensional ground-penetrating radar (3D GPR) wide-area screening, manual interpretation, two-dimensional (2D) multi-frequency verification, and borehole endoscopic validation. Typical electromagnetic response characteristics of cavity, void, and loose-zone defects were summarized, and qualitative recognition criteria for different defect types were established. A total of 153 endoscopically validated defect cases, including 115 loose zones, 28 voids, and 10 cavities, were analyzed to investigate defect distribution and associated formation factors. The results show that underground pipeline damage, engineering disturbance, and inadequate backfill compaction are the main factors associated with defect development. Inadequate backfill compaction accounts for approximately 70% of all detected defects and is mainly related to early-stage defects, whereas pipeline damage is associated with nearly 90% of cavity cases despite accounting for only 30% of all cases. Risk assessment results show that cavities and voids are mostly high-risk defects, while loose zones are mainly moderate- to low-risk defects. The proposed framework provides a practical basis for subsurface defect identification, risk warning, targeted remediation, and preventive maintenance of urban roads.