Re: The "American Korolev"
Well, any design process is only as good as the engineers following it...
Which is, in my opinion, to say that the engineers have to understand what the process is supposed to accomplish and what its rules are designed to avoid. There is no magic bullet to effective design. That's not to say formalisms and procedures are ineffective. But it's to say that formalisms and procedures do not guarantee success. Which, now that I read your post more carefully, you say eloquently...
However, if the people using the tools are clueless, the process cannot save them.
In Six Sigma’s defense, it does provide a framework for quantitatively refuting bad ideas, which beats qualitative approaches any day.
Yes, and if you read von Braun's response to LOR, for example, you see initially a qualitative rejection, which von Braun himself overturned when presented with a quantitative justification. When things like fuel budget, complexity, and safety are quantified, LOR made sense. Von Braun recognized this. He wasn't "overruled". He just waited for a more persuasive argument.
I wish -- oh, how I wish -- more of these pie-in-the-sky ideas were subjected to more rigorous examination before they generated so much excitement. Every single design process I know about starts with a "blue sky" session where any solution is invited. But the goal of these procedures is to remove irrational sources of objection: the "oh, that'll never work" mode of rejection. It isn't mean to suggest that all design proposals have equal merit.
In my experience, it merely makes the manager over-reliant on advice from senior engineers that is often skewed towards a particular solution or process. Worse, it leave the manager incapable of telling when a program is out of control or headed in the wrong direction.
In my experience these outcomes invariably follow the appointment of an inexpert manager.
Managers don’t have to be engineers to know a web board from my real work.
Maybe we can get BA to redo the site so it looks more like Pro/ENGINEER or SolidWorks. :-)
|