Document Type : Review Article
Authors
1
Master's degree, Faculty of Civil and Environmental Engineering, Amirkabir University of Technology, Tehran, Iran
2
Postdoctoral, Faculty of Civil and Environmental Engineering, Amirkabir University of Technology, Tehran, Iran, rashidtanzadeh@aut.ac.ir
3
Dep. of Civil Engineering, Amirkabir University of Technology
Abstract
Incorporating deficient design and construction methods in the pavement industry, exacerbated by unoptimized maintenance plans, has led to unprecedented economic and social costs. Therefore, novel technologies and up-to-date science are urgently needed. Artificial Intelligence (AI), one such technology, is used to develop machines and algorithms that mimic the human brain. AI has proven cost- and time-effective in enhancing asphalt pavement design, material production, construction, and maintenance management, compared to traditional solutions. This article delves into the applications of Machine Learning (ML), a subset of AI, in pavement engineering by reviewing 150 related scientific articles. The results show ML has been employed in seven research areas: design optimization (11% of studies), asphalt performance prediction (8%), prediction of asphalt mixture characteristics (33%), detection of surface defects (19%), classification of surface defects (2%), prediction of pavement functional indices (21%), and maintenance plans optimization (6%). Statistical analyses on publication frequency, algorithms used, and input features for predicting performance were presented to outline trends, research gaps, and achievements. It is concluded that ML is an indispensable tool for improving, optimizing, and conducting critical processes in pavement design, material production, construction, and management. Consequently, further research into ML applications in pavement engineering is necessary. This will facilitate the development of cutting-edge technologies like Digital Twins (DTs) for the industry.
Keywords
Main Subjects
)