A Digital Twin Framework for Traffic-Aware UAV Pavement Monitoring without Lane Closure

Abstract

UAV-based pavement inspection can reduce the cost and risk of road-surface monitoring, but real-world deployment remains difficult when traffic, pedestrians, and temporary occlusions affect the visibility of defects. This paper presents a Unity-based digital twin framework for traffic-aware UAV pavement monitoring without lane closure. The proposed environment integrates procedurally generated road defects, dynamic vehicles and pedestrians, autonomous UAV navigation, and an embedded road-damage perception pipeline. The perception module uses a two-stage approach: a lightweight YOLOv8n detector first localises road defects, pedestrians, and vehicles, while a second classifier distinguishes among potholes, single cracks, and crocodile cracks. On the simulator test set, the full pipeline achieved 99.26% overall accuracy across five classes. The digital twin was then used to evaluate three recovery strategies for occluded road segments: hover-and-recheck, micro-repositioning, and skip-and-revisit. Experiments were conducted across different traffic densities and flight altitudes using coverage, mission time, energy consumption, and revisit ratio as operational metrics. Results show that flight altitude has a strong influence on inspection coverage and that adaptive recovery improves performance under occlusion. In particular, hover-and-recheck achieved the most consistent coverage under medium and high traffic conditions, reaching up to 97.03% coverage, while skip-and-revisit was most effective in low-traffic scenarios, reaching 97.95\% coverage at medium altitude. These results demonstrate that digital twins can support the development and evaluation of traffic-aware UAV inspection strategies before real-world deployment.