I know comparatively little about general engineering methodologies, but it strikes me that the (now outdated) "waterfall" method is possibly very apt for mechanical engineering. Perhaps advances in rapidly changing software engineering could also be applied.
Are software methodologies influencing other engineering disciplines? Or vice versa?
Wikipedia's page on engineering says:
Koen argues that the definition of what makes one an engineer should not be based on what he produces, but rather how he goes about it.
Which vaguely implies that development processes are to some degree ad-hoc, given that neither the methodology section of that page, nor the problem solving subsection of the engineering page make any references to development stages in the precise ways that is widely encouraged in software engineering.
So why the apparent lack of processes?