Measuring pavement maintenance effectiveness using Markov Chains analysis

Abstract

This study assesses the effectiveness of road network pavement maintenance using Markov Chain analysis based on historical costs and road roughness progression data. The analysis is based on a database developed by the State of Victoria, Australia, consisting of 2197 road sections. The analysis measures maintenance effectiveness using the criterion of whether road sections remain in the same condition state or move to the next worst state based on a predefined roughness value. Principal inputs for the stochastic models, such as the development of transition probability matrices and associated cost functions, are discussed. Results show that, within the current budget envelope and when undertaking only routine maintenance, the probability of road sections remaining in the same condition state, which is a determinant of maintenance effectiveness, exhibits a declining tendency from good to worst condition states. This finding prompts the discussion on when to begin intervention using high types of maintenance together with their respective higher expenditure levels. The method discussed in this paper, provides tool for road authorities to select the appropriate maintenance action for each condition state of pavements based on the comparison analysis of the results of Markov Chain for different types of maintenance actions.