Development of a Cognitive Digital Twin for Pavement Infrastructure Health Monitoring

Abstract

A road network is the key foundation of any nation’s critical infrastructure. Pavements represent one of the longest-living structures, having a post-construction life of 20–40 years. Currently, most attempts at maintaining and repairing these structures are performed in a reactive and traditional fashion. Recent advances in technology and research have proposed the implementation of costly measures and time-intensive techniques. This research presents a novel automated approach to develop a cognitive twin of a pavement structure by implementing advanced modelling and machine learning techniques from unmanned aerial vehicle (e.g., drone) acquired data. The research established how the twin is initially developed and subsequently capable of detecting current damage on the pavement structure. The proposed method is also compared to the traditional approach of evaluating pavement condition as well as the more advanced method of employing a specialized diagnosis vehicle. This study demonstrated an efficiency enhancement of maintaining pavement infrastructure.