Wow...for want of solder, endurance was lost. A nice way of outlining it.
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>
Along those same lines, how do other teams control quality, short of watching over every single thing done on the car? <HR></BLOCKQUOTE>
Cornell's generally done it more or less informally. While valuable and all, all the TQM or Six Sigma rhetoric in the world cannot instill quality as a cultural value, and while many of the methods used in these approaches can reduce failure considerably, the overhead is usually insane and well beyond the ability or interest of a FSAE Team.
But Cornell's team is certainly quality-driven, by cultural expectation and peer pressure for the most part. Experienced folk know the high standards we've met in the past and try to instill that ethic to newer people. As a result, when one looks at the car getting finalized in April, most people feel compelled to bring your part or system up to the caliber of others.
Of course, quality can also be (and usually is) a vicious cycle. If you see somewhat sketchy systems on the car, you'll be more likely to slack off on your own, since hey, they got away with it, right?
It's similar to dealing with litter. The moment people see litter accumulate without being picked up, the more likely they'll litter themselves.
Conversely, it's more difficult to litter in a very clean environment. And if you do it anyway, others around you are more likely to clean up after you (and tell you off...)
It's a subtle but very real problem, and the solution to litter and car sketchiness is the same - identify and eradicate sketchiness early to avoid a low standard being set.
One simple way Cornell's done this in the past is establishing a "sketchy list". All team members vote on what they believe are the five most sketchy parts/systems. Inevitably, you get about 40-50 items, with three or four universally regarded as huge problems.
The important and obvious problems are usually tackled quickly. If it means shuffling people around and even taking people off a project, well, so be it.
The act of doing this seems to spur others into cleaning up their bits, however minor - even refining and tightening up parts that aren't on the list. In a really solid year, there's few if any items that haven't been addressed.
Of course, having the car done early and having the time to break it repeatedly helps a great deal in all of this. Few parts on the Feb. 1 car will be 100% the same May 15th.
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>
Does anyone have a back up? I mean do you have another guy who knows about a particular part of the car in case something happens to guy who designed/made it?
<HR></BLOCKQUOTE>
It's a great idea if you can do it. When one person establishes exclusive knowledge in a domain, the more irreplaceable he/she becomes. And it's not only a problem if they get overstressed, sick, have personal issues, etc. - when they graduate, you can get really screwed.
The formal term in social network analysis for this is "structural hole" - a node in a network that has numerous non-redundant ties, such that the node's removal creates significant disruption.
Indispensible people are great to have around, but it's a good idea to have someone shadow them, if only to learn from them and retain that knowledge for future teams.