Road Performance Prediction Model for the Libyan Road Network Depending on Experts’ Knowledge and Current Road Condition Using Bayes Linear Regression

Abstract

The accurate prediction of rates of road deterioration is important in Pavement Management Systems (PMS), to ensure efficient and forward looking management and for setting present and future budget requirements. Libyan roads face increasing damage from the lack of regular maintenance. This reinforces the need to develop a system to predict road deterioration in order to determine optimal pavement intervention strategies (OIS). In a PMS, pavement deterioration can be modeled deterministically or probabilistically. This paper proposes a Bayesian linear regression method to develop a performance model in the absence of historical data; instead, the model uses expert knowledge as a prior distribution. As such, Libyan Road experts who have worked for a long time with the Libyan Road and Transportation Agency have been interviewed to develop input data to feed the Bayesian Model. A posterior distribution was computed using a likelihood function depending on road condition inspections in accordance with a pre-established protocol. The results were the pavement deterioration prediction models based on expert knowledge and a few on-site inspections.