Estimation of dynamic performance models for transportation infrastructure using panel data

We present state-space specifications of time series models as a framework to formulate dynamic performance models for transportation facilities, and to estimate them using panel data sets. The framework provides a flexible and rigorous approach to simultaneously capture the effect of serial dependence and of exogenous factors, while controlling for individual heterogeneity when pooling data across the facilities that comprise the panel. Because the information contained in time series and cross-section data are combined in the estimation, the ensuing performance models capture effects that are not identifiable in either pure time series or pure cross-section data. Also, pooling data across facilities leads to improved estimation results. To illustrate the methodology, we consider three classes of models for a panel of asphalt pavements from the AASHO Road Test. The models differ in the assumptions regarding the structure of the underlying mechanisms generating the data sequences. The results indicate that serial dependence is indeed significant, thereby reinforcing the importance of dynamic modeling. We also compare the specifications to assess the poolability of pavement condition data. The results provide evidence that heterogeneity among the facilities is present in the panel. Finally, we highlight features that elude existing performance models developed with static modeling approaches: the ability to estimate maintenance activities as exogenous variables, and the capability of updating forecasts in response to inspections.