Relating pavement distress to serviceability and performance

Regression analyses, Markov processes, Bayesian analysis, and utility theory were examined as analytical tools useful in relating distress to performance. Several specific performance models of limited applicability are reported, along with some generally applicable techniques and recommended methods. It is shown that useful relationships can be obtained from existing data and are, in fact, being used in several states and other highway agencies. In each case, however, it is shown that significant improvements in the distress performance relationships are needed which can only be obtained through the conscientious observation of distress and performance of pavements under a variety of environmental and load conditions for a reasonable period of time, e.g., (approximately ten years).