If architecture is about the fundamental or unifying structure of a system, architecting must be about setting, defining, or describing that structure. This, by itself, does not distinguish architecting from other systems engineering activities. What distinguishes architecting is the need to define and understand the specific realms of practice. To help visualize the relationships among these realms, Aerospace developed the so-called "footprint" table, which describes them according to various characteristics, as shown below. For example, the footprint for a "textbook engineering" case would show that the sponsors and users are obvious and available, the technology is not a challenge, and the problem is clearly separated from the solution. A "classic" systems engineering case would be modestly more complex: In this case, the technology level is typically higher (and thus riskier), finding feasible solutions requires sophistication in design space enumeration and optimization, and the essential objectives of the underlying problem are no longer presented directly but must be discovered through elicitation and analysis.
Systems architecting deals with design problems that are even more complex—problems in which the situation and objectives are ill-structured and the quality is semimeasurable. An ill-structured problem is one in which the statement of the problem depends on the statement of the solution; clearly, this characterizes many real design situations. Although designers often act as if all requirements issues can be resolved by better elicitation, experience shows that they cannot, even in relatively familiar environments. In reality, users change their own understandings of needs and priorities when presented with design alternatives. Semimeasurable quality implies that the desired quality levels can be quantitatively described, but not directly measured. For example, the desired probability of launch failure is zero, but usually specified as a number in the range of 10-4 to 10-6. Such numbers can never be directly verified because only a handful of vehicles are ever launched. Thus, the analysis of semimeasurable quality is always a mixture of analytical and heuristic methods—analytical methods to verify that the models are accurate, and heuristic methods to ensure that they are adequate.
| Simple | Complex | |||
|---|---|---|---|---|
| Sponsors | One, w/ $ | Several, w/ $ | One, w/o $ | Many, w/o $ |
| Users | Same as sponsors | Aligned with sponsor | Distinct from sponsor | Unknown |
| Technology | Low | Medium | High | Superhigh |
| Feasibility | Easy | Barely | No | |
| Control | Centralized | Distributed | Virtual | |
| Situation- Objectives | Tame | Discoverable | Ill-structured | Wicked |
| Quality | Measurable | Semimeasurable | One-shot and unstable | |
Textbook engineering footprint.