6.0 ELEMENTS OF MODELING
J.F. Traub explores the computational complexity and consequent intractability of some discrete and continuous problems within the computational model. The ties between computation and formalized scientific truth is clear. Framed within the limitations of computing technology, particularly in terms of processing speed, it is argued that there exist some scientific problems that are computationally intractable and thus unknowable [Traub 90].
Measures of complexity form the basis of cost estimation throughout all domains of business and engineering. Complexity measures span a number of orthogonal metrics to predict project time as well as determine resource requirements. Software engineering attempts to impose the rigorous discipline of other engineering fields and the specific attempts to measure complexity within a software project can be applied to engineering modeling in general.
Software as an artifact provides a common, testable base in which to research design. One of the primary elements of software design is the program; virtually all the overhead of planning and conceptualization is done on paper and in text. Currently, software programs can more easily be parsed and analyzed than the representations of other engineering designs, such as CAD models. As such, we look toward software complexity and computational complexity for help in determining the issue and metrics required for an overall examination of engineering design complexity.