Impact of Error in Pavement Condition Data on the Output of Network-Level Pavement Management Systems

Abstract

The quality of pavement condition data is important not only in assessing the current condition of the network but also in predicting its future condition and planning future maintenance and rehabilitation (M&R) activities. This paper provides a quantitative assessment of the impact of error magnitude and type (systematic and random) in pavement condition data on the accuracy of pavement management system (PMS) outputs (i.e., forecasted needed budget and M&R activities in a multiyear planning period). The process developed to simulate the propagation of pavement condition errors to PMS output consisted of five components: condition data generation, error perturbation, condition prediction, M&R prioritization, and output generation. This process was applied to the 2011 pavement condition data set of the Bryan District of the Texas Department of Transportation. In 2011, the roadway network of the Bryan District consisted of approximately 3,200 roadbed centerline miles. The study results show that both systematic and random errors can highly distort some PMS output parameters, even in error ranges that may be considered acceptable in practice. These effects tend to persist throughout the planning period. The findings of this study can help highway agencies optimize processes for collecting pavement condition data by focusing on error levels and types that cause the greatest impact on PMS output.