System Integration and Test
During the design and construction of complicated systems, one usually attempts to build and test the components of the system individually, then to hook these components together to build up the complete system. This process is called system integration and it is very important that adequate provision be made for it in the budget and schedule of an engineering program.
Once an operational subset of the system is assembled, it is tested against its specifications. Depending on the quality of the system engineering and of the design effort that went into the project, it is at this point that a multitude of unexpected problems may appear. If they do appear, it is often a crisis, because the deadline for delivery of the system is usually near, if not already past.
In electro-optical systems, common causes of problems that appear in the system integration phase are: electronic noise and crosstalk, stray light, and various problems due to vibration such as microphonics.
During testing, puzzling results are not uncommon, even when system is working correctly, as designed. Even with the best engineering, it is difficult to anticipate everything; and often schedule and budget pressures restrict the engineering effort. As a result, the designed and specified tests are sometimes inadequate to ensure that the system is working fully, thus one must be on the lookout for puzzling and/or seemingly inconsistent results because these are important clues that something may be amiss. Ad hoc “sanity checks” are often useful in clarifying such situations.
I have integrated and tested the same sorts of systems that I have designed. System integration takes the same sort of skills as does system design, but here the approach is generally much less formal. Quick and dirty stopgap solutions are often just what are needed when schedule and budget pressures dominate.
Of course, when one is working with simple systems, integration and test are not formally separated from design and construction and are handled as a matter of course.
|
Copyright © 2002, David F. Schaack. All Rights Reserved. |