Reliable cost estimation of systems in the early design stages can be very beneficial. Some methods that do focus on system estimation are based on high level parametric models, which are not always detailed enough to provide a reliable or useful prediction model.
This thesis provides a repeatable method for creating Cost Estimating Relationships (CERs) for the estimation of a system of parts using attributes available at a preliminary design stage. The methodology provides guidelines for collecting data, classifying data, and creating CERs. Also, Attribute Estimating Relationships (AERs) are discussed for estimating the CER inputs that may not be available in preliminary design with attributes that are available in preliminary design.
The methodology was tested on a system of parts from a jet engine and the resulting CERs were compared to other higher level models. The results indicated the proposed method was slightly less accurate than some alternative models. However, the proposed method provided a more detailed and logical CER than the alternate models.