As an alternative to the mechanistic based prediction models it is possible to set up pure empirical relations as a function of age of pavement or accumulated traffic. Those may be used where structural evaluation has not been carried out, or they may be used in combination with the mechanistic models to ensure a minimum decrease in performance as a function of deterioration not related to the structural condition.

Equation 2: Example of empirical prediction model.

The models are in general expressed as seen in equation 2, where traffic (N) in this case is used as the depending parameter.

The constants IRI0, A and B can be calibrated for each section.

Facilities exist for automatically calibrating the models, based on existing measurements, and allowable ranges for the constants given by the user.

Similar models can be entered for prediction of rut depth, surface wear and friction.

In case no existing measurements are present for a section, the models can be automatically calibrated using general input for the initial and terminal condition.