Search This Blog

Tuesday, November 30, 2010

Requirements Reuse and Decision Management

Our objective for PBSE is to achieve reuse of more than diagrams, documents and models; these items are part of defining requirements and reuse of requirements is also an objective. Decisions are a central part of the requirements analysis process and reuse of decisions facilitates reusing requirements. Reuse saves time and money but it is not achieved for free. Requirements and decisions must be managed properly in order to achieve reuse. Understanding the importance of decisions in requirements analysis is helped by examining requirements analysis via the diagram in Figure 5-1. The ideas presented here are reexamined in the next chapter but this preview is provided that helps explains why so much attention is paid to the details of requirements analysis in the next chapter.


Figure 5-1 Decisions are an essential part of requirements analysis.

It is important to recognize that the value in reusable decisions is for more than facilitating the reuse of a previously used requirement or to facilitate reuse on a subsequent system development. Decisions are reusable during the life cycle of products. Problems can arise in system test or in production whose resolution requires understanding the rational for the design decisions. Potential design changes may be considered for cost savings or due to obsolete or unavailable components. In all of these cases if the rational for design decisions is not readily available and trusted then systems engineering work has to be repeated. Thus the extra cost incurred in making decisions reusable is more than recovered through savings from both reuse and from avoidance of rework.
The process of making decisions involves defining criteria, developing alternatives, conducting trades among the alternatives using a formal or informal trade study process that considers other input data to the requirements analysis process, e.g. technology readiness levels and supplier capabilities, and selecting the best alternative for the defined criteria. The NASA Systems Engineering Handbook has extensive discussion of decision analysis that is well worth reading. Two useful tools included in the NASA description are the decision matrix and an outline for a decision report. A decision matrix is a useful tool for making decisions based on weighted criteria.
Requirements from customers or market analysis drive criteria that are used by systems engineers in making decisions about system requirements. The resulting system requirements drive criteria for making decisions about lower level requirements. In describing requirements analysis verbs such as allocate, flow down, and estimate are used but all of these are fundamentally a decision making process as shown in Figure 5-1.

Figure 5-1 shows how closely requirements and decisions are connected and therefore why the desire for both reusable decisions and reusable requirements. If requirements are reusable time is saved but if the related decisions are reusable then design attributes, risks and issues driven by the decisions are understood. If the decisions are reusable then it is likely that the designs attributes are reusable; risks created by the decisions are likely to have been mitigated in previous work and issues created are likely to be known and resolved.
Achieving reusable requirements and decisions involves additional work but work that pays dividends both during the system life cycle and in future developments. Figure 5-2 shows that requirements traceability is the foundation of sound requirements analysis but traceability alone is not sufficient for either sound systems engineering practice or for achieving reusability. The attributes of requirements resulting from a robust requirements analysis process that leads to control and then to reusability are indicated in the figure.

 Figure 5-2 Achieving requirements reuse requires comprehensive requirements analysis.

Similarly achieving reuse of decisions also requires a comprehensive process as shown in Figure 5-3. The use of the pyramid format in the figures is meant to indicate that there are benefits from the doing the additional work necessary to move up the pyramid. It will be helpful for the reader to review these two figures after reading the sections on requirements analysis in chapter 6.


Figure 5-3 Planning and analysis is necessary to achieve reuse of decisions made in requirements analysis.

The close connection between requirements management and decision management suggests the two can be integrated into a common process called data management. Data management is facilitated if the tool used for requirements management has the capability for including decision modules or links to technical memos, as do some modern requirements management tools. If simple tools like spreadsheets are used for requirements management for small systems then inset links to technical memos that capture the decision rationale and supporting data associated with requirements. Properly capturing decision rationale and supporting data helps eliminate rework and facilitates reuse of requirements analysis work. The outline for a decision report described in the NASA Systems Engineering Handbook is a good guideline for properly documenting decisions and making decisions reusable.

No comments:

Post a Comment