Data-Driven Pavement Performance Modelling: A Short Review

The road network in many countries holds significant value, and the amount will continue to grow due to the rapid growth of cities and populations. Given this, it is crucial to improve knowledge of the existing road assets, based on which effective management strategies can be customized to maximize the useful service life of current road assets. A key aspect in achieving this goal is the employment of performance modelling, which forecasts future pavement performance. In recent years, the increasing popularity of data-driven approaches has propelled the development of advanced pavement performance models. In this paper, existing data-driven performance models developed globally for different types of pavements and various climate and environmental conditions are first summarized. The review on data-driven performance models then focuses on the capabilities of these models: (i) in handling the time-dependent nature of the data involved, and (ii) in utilizing the existing information available to engineers to forecast future pavement conditions. The objective of this review is to highlight the current state-of-the-art and challenges in data-driven performance modelling and conclude with potential directions and insights for driving innovation and research in the roads sector for practical applications.