Database development for an HMA pavement performance analysis system

55 p.The primary purpose of this report was to develop a database template, using the existing Wisconsin Department of Transportation (DOT) pavement management system, from which to perform pavement performance analysis using design, construction, and performance data for hot-mix asphalt (HMA) pavements. A second purpose was to investigate appropriate numerical or statistical methods that have the potential of quantifying and establishing relationships between design, construction, and performance data. A series of tasks was conducted including a review of literature, review of Wisconsin DOT databases, database integration with emphasis on performance modeling, and recommended approaches for performance modeling. The literature review found that data types collected for performance evaluation and modeling vary among agencies depending on needs, but the most common types include inventory, condition, traffic volume, and maintenance and rehabilitation. Common referencing systems between various data collection systems can facilitate data integration for pavement performance modeling; however, a major barrier for achieving full data integration is lack of common referencing systems compounded by the use of different data formats. To that end, Geographic Information System (GIS) was identified as an effective tool for data integration among various divisions within an organization. Several Wisconsin DOT databases applicable to performance modeling for HMA pavements were reviewed for primary data categories including construction, design, traffic, and performance. Semantic discrepancies among databases that impede integration were summarized, then recommendations were identified to enable simple or complex queries to relate data residing in the different databases. A GIS-based database integration was recommended using similar Wisconsin DOT GIS practices. A loose coupling approach, involving the transfer of data files between the GIS and other programs, was demonstrated using screen snapshots. Then, the integrated data were prepared for export into a statistical analysis package and the results imported back to the GIS for data visualization or display. Several statistical analysis methods to develop performance models were provided, along with reference examples for ANOVA (analysis of variance), comparison of means, and regression models. Currently there is an on-going research study with estimated completion in 2008, NCHRP Project 9-22, Beta Testing and Validation of HMA PRS, that will develop software capable of developing pavement performance models