Abstract
The United States is facing a challenge on repairing aging civil infrastructures because of limited resources allocation. The international roughness index (IRI) is a key quantity of road inspection data to aid officials in prioritizing road maintenance decisions, which allows them to be efficient with their resources. IRI measures both the road roughness and riders’ comfort level. This paper introduces a probabilistic method that analyzes probability density function (PDF) of acoustic data collected from a vehicle-mounted microphone to estimate IRI. An analysis of noise source is conducted to support the accuracy of the approach. Compared to existing approaches, the present method is more economical, easier to install, and requires less space. Curves of probability density function (PDF), which show the probability of each sound pressure occurring over a stretch of road, are used to estimate IRI based on the principle that a probability distribution concentrates at low sound pressures over smooth roads and at high sound pressures over rough roads. This method is validated using data collected by a test vehicle that was driven around an engineered test track. Primary conclusions are: (1) the Weibull distribution performs better than normal and lognormal distributions; (2) the optimal Weibull distribution parameter is identified; and (3) the method performs well when applied to different pavement designs, namely, Superpave, Stone Matrix Asphalt (SMA), and Open Grade Friction Coarse (OGFC).
