GPR monitoring for road transport infrastructure: A systematic review and machine learning insights

Abstract

Suitable road pavements assessment becomes essential to provide safe traffic movements of people and goods. Moreover, a reliable transportation network is a crucial aspect of economic growth. Road pavements are subjected to various factors that influence overall performance (e.g., traffic load, temperature, moisture, delamination of the pavement layers, subsurface condition, etc.). These factors can reduce the infrastructure's life and decrease the circulation comfort of the vehicles in the transportation network. Early inspection of pavements optimizes maintenance and repairing methodologies, decreasing the maintenance cost and increasing the lifespan of the road pavements. Non-destructive techniques are strongly recommended to achieve accurate and valuable information from the subsurface condition. Ground Penetrating Radar (GPR) is a non-destructive geophysical method widely used on infrastructure assessment, particularly in road pavements, due to its low operation cost, time-saving, non-invasive, and less workforce. This paper presents a critical state of the art of applying GPR to diagnose road pavement and detect inner damages such as debonding, sinkholes, moisture, etc. The incorporation of the GPR with other complementary techniques in pavement inspection is also discussed. Through the review, the GPR capabilities for road inspection and evaluation of subsurface identification have been successfully demonstrated and validated in numerous studies and case studies. Finally, the application of more recent processing techniques to support decision-making owners/operators, such as machine learning and intelligent data analysis methods, and the future challenges on the GPR application in road pavements are introduced.