Capstone Project Wiki

Poor artifacts result in poor design transferability, which invariably reduces the desirability of the product. Thus making appropriate, high-quality artifacts in a timely manner is essential to the design process.

We note that many students see artifact creation as merely “paper” work, an attitude that tends to result in poorly created artifacts, made primarily as an afterthought, that do not meet the needs of the sponsor. The following considers three major times when artifacts should be made.

One strategy for creating artifacts is to wait for the end of a stage of development, then prepare the artifacts that are necessary for obtaining stage approval. This has the advantage that the reason for each artifact is clear. It has disadvantages including: a) the fact that significant time will have passed since key decisions were made, so artifact details may be somewhat fuzzy; b) some artifacts may be forgotten, since not all useful artifacts are required for stage approval; c) it reinforces the idea that artifact creation is separated from design work; and d) there is a tendency to just reach the minimum bar necessary for approval.

A second strategy for creating artifacts is to create them continuously. That is, as you are working on making decisions you are simultaneously creating artifacts, so that when the decision is made, the artifact is already created. This strategy has the advantage that the details of the decision are extremely fresh when the artifact is created. It might appear that this strategy is optimal, but is suffers from an important weakness: much of the design work you do will lead to dead ends. Artifacts that document dead ends may have some use, but they are not nearly as important as artifacts that document the selected design. Thus, with this strategy you are at risk of spending significant time on artifacts of relatively low importance.

The best way to balance the risks of losing details and spending time on artifacts of low importance is to create informal artifacts as entries in a design record book while you are doing the work. You record your design decisions and reasons for them in sufficient detail that you will be able to create an effective artifact later, if it's needed. You record your experimental procedures in your record book in sufficient detail to allow you to effectively repeat the work if needed. You record your test results at the time you run the tests, and you even do some preliminary analysis in your record book to see how good the results are. That way, you have captured all the information, and have it ready to formalize if you need to transfer it as part of the finished design. Your design record book is the place to continuously record your work.

A third strategy, and the one Capstone recommends, is to create appropriate artifacts when decisions requiring artifacts should be made. Thus, for mundane decisions, as soon as the decision is made, it is recorded in a drawing or bill of materials entry. When a major decision is made, in addition to capturing the decision in a design definition artifacts, an artifact providing the justification for the decision is created. For a critical decision, the mathematical analysis or experimental procedure used to optimize the design is captured in a test artifact; the interpretation of the results of the test is captured in a design justification artifact; and the decision itself is captured in the design artifacts.

This strategy has several advantages. First, the artifacts are created while the decision is still fresh. Second, minimal effort is spent creating artifacts for dead-end decisions. Third, the artifacts show a clear sense of progress in the design and are useful to the team and management in tracking the progress of the project. Finally, relatively little work is necessary to prepare for stage approval. When the design for the stage is complete, the artifacts are already complete.

Some consider a weakness of this strategy to be that it slows down progress on the design as time is taken to create artifacts. However, we believe that creating artifacts at the appropriate time actually increases the progress on the design since it can inform the process. The careful thought required by the creation of effective artifacts helps ensure that good decisions are made. Furthermore, the checking of design artifacts results in checking of the design decisions, so mistakes or weaknesses are more likely to be caught early on, which minimizes backtracking and later rework.