next up previous
Next: The Characters Up: Domain Processes and Engineering Previous: Domain Processes and Engineering

Introduction

The objective of this working group was to identify key patterns and choices that can impact a Domain Engineering effort for better or worse, and characterize these patterns/choices at a more specific, lower level of granularity than is obtained with conventional process description languages. Of particular interest was to identify patterns of interaction among stakeholders that could postively or adversely impact the effort.

Rather than using the traditional techniques such as brainstorming, cause/effect diagrams (also known as fishbone or Ishikawa diagrams), affinity diagrams (also known as the KJ method), etc., we adopted a novel approach using the metaphors of theater. This method attempts to uncover and analyze some of the key "scenes" in the theatrical drama of domain engineering.

Although this technique probably derives information similar to the traditional techniques, the participants have more control in understanding the sequence of events that brought us to the conclusions. Most of the traditional methods produce random lists of information, which are then organized on the basis of undocumented criteria. The "theatrical script" technique provided for a logical sequence of documented events that led to the conclusions, as well as the identification of key decision points (or issues) that may have altered the final conclusions. This technique does require more effort than the traditional techniques, to traverse and document the alternative paths in a tree of script scenarios .

In this report, we present the resulting drama in terms of its characters, their relationships in the organization, and a succession of scenes that represent one path for a domain engineering effort. This particular path represents a series of choices that are close to the borderline between success and failure. We did this in order to surface as many issues as possible. Interactions between characters that indicate significant risks are flagged with an asterisk (*), and for each such risk we identify potential alternative responses that might increase the chances of success.



next up previous
Next: The Characters Up: Domain Processes and Engineering Previous: Domain Processes and Engineering



Larry Latour
Tue Sep 19 00:30:00 EDT 1995