ABSTRACT
The prediction of road behavior is characterized as a variable process defined by several exogenous factors, such as traffic loads or climate, and endogenous factors, for instance planning, type and quality of construction. This paper discusses the procedure of analyzing big data and potential sources of error by using a pure statistical approach. Therefore, explanatory variables such as age of surface layer, traffic loads and structure thickness were analyzed concerning their power of influence on rut depths. This investigation was based on a part of the Bavarian road network. The results confirmed that the age of a road section, structure thickness and traffic load are effective variables to explain the prospective condition of roads. The used methodological approach offers substantial advantages but has drawbacks as well that have to be taken into account and which are also discussed.
