Comprehensive simulation for end-result specification development

End-Result Specifications (ERS) for asphalt pavement construction offer potential benefits over method-related specifications. ERS in general, has led to the improvement in quality of the constructed pavements. Performance related specifications (PRS) are more rational and more desirable because life cycle cost analysis can be performed. The present lack of accurate deterioration models in the pavement field, however, do not allow for the development of a true PRS. While ERS systems for asphalt pavement construction represent the current state-of-the-art, the ability to assess quality and assign appropriate payment adjustment is an imperfect system, or otherwise stated, it is an enterprise with a significant associated risk that must be addressed. In order to properly utilize and administer a contractual process involving risk, one must first have an accurate measure of that risk.; This work details the development of a simulation tool which can be used to analyze specification risk and to develop ERS systems with user-managed risk levels. Modeling and simulation of Illinois ERS highway project data clearly demonstrated that the magnitude and relative amounts of production and measurement variability play a key role in ERS risk. A novel procedure has been presented for estimating variabilities based on sound statistical principles. A Composite Risk Index (CRI) has been proposed to provide a single parameter for quantification of payment risk. An analysis of the old and new MOT pay factor equation for Superpave asphalt ERS was conducted, which highlighted the pros and cons of the new pay factor formula. Specific examples of how IDOT's existing ERS for Superpave HMA could be modified to reduce contractor and agency testing are also presented. An attempt has been made to define a strategy for developing a surrogate simulation model using a combination of tools like neural networks, decision trees and support vector machines.