Set the quality standards for a product during the requirements stage. Treat the desired quality level as a requirement. Weigh the merits and costs of various options you have for applying quality measures to the project. Some of the trade-offs to consider include ease of use versus power and complexity, and speed versus robustness.
For short projects, used only in-house as tools or quick prototypes, you do not need to emphasize robustness. For example, if you decide to develop a VI to benchmark I/O and graphing speeds, error checking is not as crucial.
However, with more complicated projects that must be reliable, such as applications for monitoring and controlling a factory process, the software must be capable of handling invalid inputs. For example, if an operator mistakenly selects invalid voltage or current settings, the application must handle them appropriately. Institute as many safeguards as possible to prevent problems. Select a life cycle development model that helps you find problems as early as possible and allows time for formal reviews and thorough testing.