Investigating the relationship between pavement roughness and free flow speeds (FFS) of roadways for FFS to be used as a network level pavement performance evaluation metric

Abstract

Road agencies require regularly updated pavement condition data for effective maintenance planning and network-level asset management. The International Roughness Index (IRI) is widely used to evaluate pavement roughness; however, conventional roughness surveys require dedicated equipment and are often costly to repeat frequently over large networks. This study investigated whether 85th percentile Free Flow Speed (V85) can be used as a supplementary indicator for network-level pavement performance screening. IRI data were collected using the RoadLab Pro smartphone application, while speed data were extracted using the Google Maps Distance Matrix API. Five selected two-lane road sections were analyzed at 100 m, 200 m, and 500 m segment lengths. Linear regression was first used to explore the IRI–V85 relationship, and Kruskal–Wallis testing was then applied to determine whether V85 differed significantly between IRI categories. Most regression models showed negative slopes, indicating that speed generally reduced as IRI increased. However, R² values were weak to moderate. The pooled Kruskal–Wallis p-values were 0.300, 0.181, and 0.075 for 100 m, 200 m, and 500 m segments, respectively. Therefore, statistically significant V85 differences between IRI categories were not confirmed. The study concludes that V85 should not replace IRI but may support preliminary pavement screening. Future research can be conducted by controlling for mode mix, land use characteristics, traffic volume, and other factors that influence vehicle speed to better evaluate the relationship between V85 and pavement condition.