Machine Learning-Based Predictive Maintenance for Airport Pavement Management
Faculty Mentor Name
Aaron Jones
Format Preference
Poster
Abstract
Airport runways are critical safety assets that must remain in good condition to support safe aircraft takeoff and landing. Over time, runways deteriorate due to aircraft loading, weather exposure, and operational use, leading to surface defects that can reduce friction, drainage, and structural performance. These issues increase the risk of hydroplaning, foreign object debris, and runway excursions. Most airports currently rely on periodic visual inspections and reactive maintenance, which often identify problems only after deterioration has already progressed.
This project focuses on developing a data-driven approach to better understand and predict runway pavement performance. Pavement condition is evaluated using the Pavement Condition Index (PCI), a standardized rating system that scores pavement health on a scale from 0 to 100 based on observed surface distresses. Historical PCI data, along with environmental, structural, and operational variables, are analyzed to identify key factors contributing to runway deterioration. Machine learning and statistical modeling techniques are used to examine deterioration trends and improve the ability to forecast future pavement condition.
By shifting from reactive inspections to predictive analysis, this research aims to support earlier and more effective maintenance planning. The results provide airport operators with clearer insight into how and why runways deteriorate, allowing maintenance resources to be allocated more efficiently while improving safety and long-term pavement performance.
Machine Learning-Based Predictive Maintenance for Airport Pavement Management
Airport runways are critical safety assets that must remain in good condition to support safe aircraft takeoff and landing. Over time, runways deteriorate due to aircraft loading, weather exposure, and operational use, leading to surface defects that can reduce friction, drainage, and structural performance. These issues increase the risk of hydroplaning, foreign object debris, and runway excursions. Most airports currently rely on periodic visual inspections and reactive maintenance, which often identify problems only after deterioration has already progressed.
This project focuses on developing a data-driven approach to better understand and predict runway pavement performance. Pavement condition is evaluated using the Pavement Condition Index (PCI), a standardized rating system that scores pavement health on a scale from 0 to 100 based on observed surface distresses. Historical PCI data, along with environmental, structural, and operational variables, are analyzed to identify key factors contributing to runway deterioration. Machine learning and statistical modeling techniques are used to examine deterioration trends and improve the ability to forecast future pavement condition.
By shifting from reactive inspections to predictive analysis, this research aims to support earlier and more effective maintenance planning. The results provide airport operators with clearer insight into how and why runways deteriorate, allowing maintenance resources to be allocated more efficiently while improving safety and long-term pavement performance.