Author Description

In recent years, state highway agencies have come to understand the need for high quality pavement condition data at both the project and network levels. At the same time, agencies also realize that they have become too dependent on contractors to ensure the quality of the delivered data without any means to independently assure the quality of these delivered data. This research study therefore aims to investigate the inherent variability of the automated data collection processes and proposes guidelines for an automated data collection quality management program in Indiana. In particular, pavement roughness data (in terms of IRI) and pavement surface distress data (in terms of PCR and individual pavement surface distress ratings) are considered in this study. Quality control protocols adopted by the contractor are reviewed and compared against industry standards. A complete quality control plan is recommended to be adopted for all phases of the data collection cycle: preproject phase, data collection phase, and post-processing phase. Quality assurance of pavement condition data can be viewed in terms of (i) completeness of the delivered data for pavement management; (ii) accuracy, precision and reliability of pavement roughness data; and (iii) accuracy, precision and reliability of individual distress ratings and an aggregate pavement condition rating. An innovative two-stage approach is developed in this study to evaluate delivered data for integrity and completeness. Different techniques and performance measures that can be used to evaluate pavement roughness and pavement surface distress data quality are investigated. Causes for loss in IRI and PCR accuracy and precision are identified and statistical models are developed to relate project- and network-level IRIs and PCRs. Quality assurance procedures are then developed to allow highway agencies improve their pavement condition data collection practices and enhance applications inthe pavement management systems.