Algorithm for Estimating International Roughness Index by Response-Based Measuring Device

Abstract

Inertial profilers have been the mainstream measuring device for pavement roughness for decades, but many local government agencies still use profilographs and other simple measuring devices due to the relatively high cost of inertial profilers. Recently, researchers have shown great interest in studying smartphones and other types of inexpensive devices for roughness measurements. In this study, a portable response-based measuring device was developed, followed by a computational algorithm to enhance the analysis procedure of measured acceleration data. The algorithm includes speed normalization on acceleration data, Fourier transform, and low-pass and high-pass filtering on acceleration data in a frequency domain. The enhanced acceleration root-mean-square index (ARI) is then computed. Four test sections of 5.4-km long were surveyed by the developed device and inertial profiler mounted on the same vehicle at eight various speeds. It was found that the low-pass cutoff frequency corresponding to a 7.5-m wavelength and the high-pass cutoff frequency corresponding to a 30-m wavelength resulted in the best ARI calculation. The developed algorithm was successfully verified by an additional field survey of 169 subsections. The coefficient of determination R2 between the ARI and international roughness index (IRI) has been significantly improved from 0.502 to 0.881 after the enhancement procedure was conducted. Therefore, the pavement roughness IRI value can be reasonably estimated by an economic, portable, and reliable device. However, the potential errors due to vehicle sources should be further investigated for enhancing the practical application.