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

Researchers developed a Unity-based digital twin framework to test UAV-based pavement inspection strategies in simulated traffic conditions without requiring lane closures. The system achieved 99.26% accuracy in detecting road defects using YOLOv8n detection and classification, and identified hover-and-recheck as the most effective strategy for maintaining inspection coverage in high-traffic scenarios.

This research addresses a critical infrastructure challenge: automating road maintenance inspection while minimizing disruption to traffic flow. Traditional pavement monitoring requires lane closures or manual inspection, both costly and disruptive. The proposed digital twin framework bridges simulation and real-world deployment by creating a procedurally generated environment with dynamic traffic, pedestrians, and autonomous UAV navigation—enabling systematic evaluation before field testing.

The technical contribution centers on perception accuracy and operational strategy optimization. The two-stage detection pipeline (YOLOv8n for localization, followed by classification) achieves near-perfect accuracy on simulation data, though real-world performance typically degrades due to environmental variability, weather, and lighting conditions not fully captured in synthetic training. The comparison of three recovery strategies—hover-and-recheck, micro-repositioning, and skip-and-revisit—reveals that no single approach dominates across all conditions; instead, traffic density and altitude significantly influence which strategy maximizes coverage while minimizing mission time and energy consumption.

For infrastructure operators and municipalities, this work demonstrates substantial cost-reduction potential. UAV inspection eliminates the safety risks and economic losses associated with lane closures while accelerating the inspection cycle. The framework's modularity allows adaptation to different road types, climates, and traffic patterns, making it generalizable across regions.

Limitations include the simulation-to-reality gap and the assumption that UAVs can safely operate in current airspace regulations. Future work should validate these strategies in controlled real-world environments and integrate weather robustness.

• →Digital twin framework achieved 99.26% accuracy in detecting potholes, cracks, and vehicles using lightweight YOLOv8n detection
• →Hover-and-recheck strategy maintained 97.03% coverage in medium-to-high traffic; skip-and-revisit reached 97.95% in low-traffic scenarios
• →Flight altitude significantly influences inspection coverage and energy consumption, requiring adaptive strategy selection
• →Simulation-based evaluation enables safe testing of UAV inspection strategies before costly real-world deployment
• →Framework supports traffic-aware inspection without lane closures, reducing economic and safety costs for infrastructure maintenance