Drone-Assisted Bridge and Highway Inspection Using Computer Vision and Digital Infrastructure Management: A Comprehensive Review

Abstract

ABSTRACT Transportation infrastructure networks, comprising extensive spans of bridges and highway corridors, face rapid structural deterioration due to increasing traffic loads, environmental factors, and aging construction materials. Traditional visual inspection protocols are inherently localized, resource intensive, subject to subjective inspector bias, and present occupational safety risks. To overcome these challenges, Unmanned Aerial Vehicle (UAV) assisted inspection systems integrated with advanced Computer Vision (CV) and Digital Infrastructure Management Systems (DIMS) have emerged as a vital paradigm. This review provides a comprehensive analysis of UAV platforms, multi-spectral sensor payloads (RGB, thermal infrared, LiDAR), and deep learning computer vision architectures (including YOLOv8, Mask R-CNN, and DeepLabv3+) for concrete crack detection, surface spalling, rebar corrosion, and pavement distress monitoring. Furthermore, we examine the integration of aerial spatial data into Building Information Modeling (BIM) and Digital Twins to enable automated asset health tracking. Operational challenges, regulatory restrictions, and future research directions toward autonomous infrastructure management are thoroughly discussed.

KEYWORDS: Unmanned Aerial Vehicles (UAVs); Bridge Inspection; Highway Monitoring; Computer Vision; Deep Learning; Digital Twin; Building Information Modeling (BIM); Structural Health Monitoring (SHM).