Database preparation for pavement modeling - Virginia's experience

The successful development of deterioration prediction models requires an adequate and reliable database The task of preparing the database for developing refined deterioration models for Virginia's interstate pavements is described. At the onset, several problems challenged the modeling effort: a database containing nonhomogeneous sections unsuitable for use in model development; a user-unfriendly system incapable of efficient data manipulation; and missing and incorrect data. A computer-assisted methodology was devised to address these limitations. Following the compilation phase, the database was subjected to a three-stage screening process This screening helped identify observation points that were likely to adversely affect the subsequent modeling stage. The problems encountered during this study suggest some desirable features that pavement management systems are recommended to possess in order to facilitate prediction model development.