Time series analysis of pavement roughness condition data for use in asset management

Abstract

Roughness is a direct measure of the unevenness of a longitudinal section of road pavement. Increased roughness corresponds to decreased ride comfort and increased road user costs. Roughness is relatively inexpensive to measure. Measuring roughness progression over time enables pavement deterioration, which is the result of a complex and chaotic system of environmental and road management influences, to be monitored. This in turn enables the long term functional behaviour of a pavement network to be understood and managed. A range of approaches has been used to model roughness progression for assistance in pavement asset management. The type of modelling able to be undertaken by road agencies depends upon the frequency and extent of data collection, which are consequences of funding available. The aims of this study are to increase the understanding of unbound granular pavement performance by investigating roughness progression, and to model roughness progression to improve roughness prediction methods. The pavement management system in place within the project partner road agency and the data available to this study lend themselves to a methodology allowing roughness progression to be investigated using financial maintenance and physical condition information available for each 1km pavement segment in a 16,000km road network.