1328307 - Classification of Flexible Pavements based on Texture Data

The strong correlations between pavement texture and other surface characteristics, such as friction, drainage, and noise, are widely known. However, those relationships are not straightforward and have a strong dependency on the type and properties of the pavement surface. Incorporating surface information in prediction models has the potential to drastically increase their predictive power and aid engineers in determining which sections of the highway network are hot spots for low friction, high tire-pavement noise or high splash and spray zones. Nonetheless, in practice, pavement surface information is often unknown at the network level and must be inferred based on local expertise. The objective of this study was to address this issue by developing a sophisticated and data-driven classification model capable of identifying different asphalt pavement surfaces with a high degree of accuracy using only field texture data.
High-resolution three-dimensional texture data were collected on 21 highway sections with varying flexible pavement surface types within 60 miles from the city of Austin, Texas, using a data collection prototype developed in-house. Sophisticated data processing algorithms and meticulous feature engineering were used to ensure the data used in developing the classification model were of the highest possible quality. Based on the most recent literature on the topic, more than twenty texture summary statistics were assessed during this study to find the best combination to predict the surface characteristics of asphalt pavements.
The results were robust and promising, not only did they reveal patterns in pavement texture that can help practitioners characterize the topological features of the flexible pavement mixes in terms of texture statistics, but they also indicate that a clever use of pavement surface texture statistics is enough to develop a network level pavement surface classification model with a predictive power as high as 89 percent measured in terms of the F1 score.