Application of Patterns for Systems Engineering Process Tailoring and Reuse

Humankind's quest for coherence often provides a glimpse into the hidden order of our universe. In every domain we find patterns. We find pattern hierarchies and their dependencies and we find patterns which unite other patterns (Bateson 1979). Pattern thought can be traced as far back as Aristotle who taught that in order to become a good interpreter one must learn to recognize resemblances---which he further described in terms of causal relationships and patterns.;The pattern research addressed in this thesis is informed by the work of Christopher Alexander and others who have demonstrated the usefulness of patterns for reusing knowledge in the engineering domain. Patterns are discernable by the relationships between a certain context, a certain system of forces which occurs repeatedly in that context, and a certain configuration which allows these forces to resolve themselves (Alexander 1977).;This research provides a better understanding of how domain attributes influence process tailoring requirements and the suitability of previously tailored processes for reuse on new programs from the perspective of patterns. The research results did show that programs with similar domain attributes were subject to similar process tailoring requirements based on program characteristics and that these program defined processes could be characterized in a pattern form and reused on similar programs.;While there is much research with respect to patterns as a means of reusing corporate knowledge and engineering work products, there is little discernable progress in applying a pattern approach for reuse of systems engineering processes.;This research examined hundreds of exemplar process solutions, previously tailored for a wide range of programs, and determined that there is anecdotal evidence of process tailoring patterns based on program domain characteristics. These patterns were validate as being informative for new programs with similar domain characteristics. The fundamental question as to how similar do two (or more) systems need to be in order to consider reusing the processes that created them is only partially answered. The evidence from this research suggests that generic processes need to be sufficiently general for reuse and that pattern-based reuse must provide clear contextual instructions for where the pattern would be suitable for reuse.