Digitalization and Automation of Runway Inspection Using Unmanned Aerial Vehicles

Area of Science:

• Civil Engineering
• Geospatial Science
• Computer Science

Background:

• Traditional airport runway inspections are manual, time-consuming, and subjective.
• Accurate pavement condition assessment is critical for aviation safety and operational efficiency.
• Existing methods struggle with scalability and detailed spatial defect mapping.

Purpose of the Study:

• To develop an automated, end-to-end framework for airport runway pavement inspection and condition assessment.
• To integrate Unmanned Aerial Vehicle (UAV) data collection with deep learning and GIS for pavement analysis.
• To create a georeferenced digital pavement condition map for maintenance prioritization.

Main Methods:

• Utilized Unmanned Aerial Vehicle (UAV) for high-resolution aerial data acquisition.
• Applied deep learning-based pixel-level semantic segmentation to identify and classify pavement surface defects.
• Employed Geographic Information System (GIS) for spatial aggregation and Pavement Condition Index (PCI)-inspired assessment.

Main Results:

• Successfully detected and localized multiple safety-critical runway pavement defect types at pixel resolution.
• Generated a georeferenced digital representation of airfield pavement condition.
• Produced a runway-scale condition map demonstrating the system's capability for maintenance decision-making.

Conclusions:

• The proposed framework offers a scalable, data-driven alternative to traditional manual runway inspection.
• The system provides a practical foundation for digital condition monitoring of airport pavement infrastructure.
• Automated inspection enhances efficiency and accuracy in airfield pavement management.