Systematic pavement management requires estimates or predictions of future pavement performance so that rational comparisons may be made among alternative courses of action. Performance models are required in two distinct contexts, depending on the pavement management level involved. At the pl'ojoct level, fairly detailed and specific models are required for predicting tho performance expected for an individual pavement section. At the network level, general or average prediction models are required to provide esti· mates of the expected performance for a typical pavement or class of pavements. Accordingly, quite distinct modeling methods are indicated for these two different modeling needs. Performance-modeling requirements and data requirements for both network-level and project-level applications are discussed. An idealiied experiment to collect data for performance modeling Is presented. A specific performance prediction model based on stoahastic concepts and treating pavement deterioration as a Markov process is presented as an example of the development· of prediction models for network-level applications.

All highway agencies are f'aced with the problem of providing and maintaining a network of roadways to serve the public. This requires both a considerable capital investment and an adequate maintenance and rehabilitation program. During the past decade, various economic, social, and political factors have made it increasingly important that transportation agencies take every step to make the most beneficial use of their often-inadequate budgets. This has resulted in the rise of the pavement management system from a theoretical concept discussed by university professors to a practical reality under development and implementation throughout the nation. As a result of this increasing emphasis, both the conceptual and the practical elements of systematic pavement management have been widely discussed (.! _-11). Great strides have been made, but significant problems have also been encountered. One such problem, which will be addressed in this paper, is the difficulty in predicting pavement performance. Systematic pavement management is based on the idea that it is possible to determine, in a reasonably objective fashion, how best to use the public funds made available for providing pavements. Budgets are typically allocated on a one-or two-year cycle, and construction, maintenance, and rehabilitation activities are generally planned on an annual basis. Nevertheless, activities carried out (or postponed) now can have a significant impact on roadway conditions for several years or even decades. In order to make rational choices among alternative courses of action, it is therefore necessary to be able to predict or estimate the future performance of the roadway under each alternative action.