Development of Probability-Based Pavement Performance Curves for Pavement Management Systems

Abstract

Pavement Management Systems (PMSs) are used by transportation agencies to develop maintenance and rehabilitation programs for the pavement network under their jurisdictions. PMSs have prediction performance models to forecast the future condition of the pavement network; these models can be deterministic or probabilistic. Deterministic models are commonly used in PMSs, but they do not consider the uncertainty in forecasting pavement performance. This paper presents an approach to develop probabilistic-based pavement performance curves (PBPPCs) to address the uncertainty in pavement performance prediction. The method to develop PBPPCs is based on probability distributions fitted to pavement condition indexes (PCIs) from past history records for similar pavement type and functional class combinations. PCI percentiles at each year of the analysis period are obtained from the probability distributions to build the PBPPCs. A case study is presented using the PBPPC approach to enhance the deterministic performance model of the Metropolitan Transportation Commission Pavement Management System (MTC-PMS) also called StreetSaver®. PBPPCs are adapted to perform a needs analysis in the case study. Findings from this study reveal that there are significant differences in the budget estimates using PBPPCs when compared to a deterministic model. Hence, it is concluded that the implementation of PBPPCs will allow agencies to perform multi-scenarios sensitivity analyses in order to identify treatment and budget needs at different probability levels.