Optimization of Pavement Inspection Schedule with Traffic Demand Prediction

Abstract

In order to improve the current pavement inspection systems, many studies have attempted to produce a flexible inspection schedule by minimizing the lifecycle cost of pavement management system. However, these methods do not incorporate the risk of untimely inspection nor adopt the state-of-art intelligent transportation resources, such as traffic data from various infrastructures, flow prediction models with high accuracy, and empirical-mechanistic models for pavement deterioration process. Therefore, this paper proposes a framework to optimize a flexible inspection schedule within a risk boundary defined by pavement state prediction with traffic flow data and a more mechanistic deterioration model. The results validate the outperformance of the optimized inspection over two conventional inspection schemes-1-year and 2-year regular inspection. Also the optimized inspection is comparably more robust than the regular inspections with different traffic scenarios due to the uncertainty risk taken by regular inspections.