Spectrum - -failures are at issue An important fnference to...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
gone wrong, and presumably to learn from the -failures are at issue. An important fnference to be ._-. . mistakes, faults, and problems or others-or, at the drawn from this observation is that an efficient, very least, to feel good about not being personally - accurate failure analysis can be carried out only if the - - --- involved,At-thshAassachussttslnstite.of. .-. ..; .____ .=--.ana!y.st_has_good kn_owiedgea~~.~th-~-~y~~-~-in _.-._ Technology, In a seminar labeled “Failure of Human ...-- question; as a corollary, if he does, corrective action ._ _. Systems,” the not very surprising conclusion was drawn (among others) that a failure is often the participants as a personal rather than as a project --- failure. Thus, it is understandable that many against which to measure the event, and a judge to . individuals who do “make the papers” are not happy measure the event against the standard. We might .I When we conceived this special issue on “What depending upon the relative latitude or ambiguity of we recdgnized the possibility that. ..‘- engineers and engineering managers might be reluctant to discuss unsuccessful projects with which --* --- . they were personally.connected.:Colleagues warned would not talk for the record about failures close to -. suffice (e.g., in the case of an airplane crash). -- -- Fortunately, those dire warnings proved largely unfounded. Peter Goldmark, in the course of discussing the ill-fated EVR project (p. 95), told Spectrum staffer Rubinstein: ‘Above all; somehow observations of actual happenings).” He points out _ ;.-. I that in some cases the “fact of failure” may be based -- ,- _ ~ -.-.;.;-_;.events are significantly different in certain ways from .” some other events defined as successful arbitrarily). -- Donald Christiansen Editor ---- -- - ~ _- --. - For the purposes of this special issue. .we shall be ___ dealing mostly, but not exclusively, with hardware
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
rong? , system-type failures whose criterion for success or nonsuccess Is relatively unambiguous. Thus, we shall be less concerned with determining that a failure took place than In d‘etermlning why it took place. In such cases, the failures may be technlcal, or they may be economic. Technical failures may be related to variable characteristics (the system may not reach its design objectives, yet it still “works”), or to inoperable elements (overstressed or poor-quality components). Of. course, more sophisticated systems may contain failed elements and still function, albeit sometimes in a degraded mode. In any event, if failed devices or components are at the root of a system failure, one may set about to determine the reason why, and the mechanism by __ which the device failed. In his article on “The failure tracers” (p. 32). Roger Allan reviews the science (and art) of failure analysis. included with that article are case histories in which failure “detectives” have traced a failure to its ultimate cause. By the time you’ve finished the article, it should be clear that reliability Is not an.absolute, but a characteristic that is related to m&ion profile. The article also suggests that, while a systematic approach to failure analysis
Background image of page 2
Image of page 3
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

Page1 / 6

Spectrum - -failures are at issue An important fnference to...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online