Do designers/engineers actually test the things they draw up?

[vi edit]

The others just draw it and don't care if it takes 10 people to put the screw in.

If there is a hell I hope they spend an eternity in it having to install their horrible creations.

 

[lxskllr]

If there is a hell I hope they spend an eternity in it having to install their horrible creations.

Per the directions of outsourced Chinese technical writers :^D

 

[vi edit]

Per the directions of outsourced Chinese technical writers :^D

That's a rant deserving of it's own thread. Some of the diagrams and directions included in things are just downright punishing. It's like the pain and misery of the bastard sitting behind the desk writing it is an extension of the directions.

Somewhere in those exploded parts diagrams and wall of text is a "somebody kill me...pretty pretty please! this job sucks!" slipped in.

 

[Howard]

In my company, my internal designs are subject to testing before being released to production, and the budget is not as tight as it is in other money-squeezing companies, fortunately.

 

[Schadenfroh]

There is no test like deployment.

 

[Chriscross3234]

I'm still in school, but we've been "taught" that most companies design processes include a verification and validation phase, especially in the biomedical device industry. I've talked to several people in industry who can support this claim, but I can let you know first hand when I graduate in a year .

 

[vi edit]

I'm still in school, but we've been "taught" that most companies design processes include a verification and validation phase, especially in the biomedical device industry.

Biomed devices are a whole different ballgame. Anything hooked up to a patient has to get approved by the FDA.

 

[Gibson486]

I'm still in school, but we've been "taught" that most companies design processes include a verification and validation phase, especially in the biomedical device industry. I've talked to several people in industry who can support this claim, but I can let you know first hand when I graduate in a year .

verification and validation is done to ensure the product works as designed. It's done to find bugs, not to find faults in design (although they go hand and hand at times). Aside from electronics and software, it's usually not deployed in other industries.

 

[Modelworks]

verification and validation is done to ensure the product works as designed. It's done to find bugs, not to find faults in design (although they go hand and hand at times). Aside from electronics and software, it's usually not deployed in other industries.

Or if a company does test a design for how easy it is to work with in the real world they use the same group over and over who are already skilled at the task. Big difference between asking someone to assemble something who has been doing that sort of thing for years vs the consumer who pulls it out of the box and never has done it before.

 

[TecHNooB]

There is no test like deployment.

 

[Kirby]

If it compiles, ship it.

 

[TecHNooB]

If it compiles, ship it.

:sneaky::whiste: D:

 

[IcePickFreak]

I'm with the others citing "cost cutting" by upper management. I actually had the GM at the last place I worked tell me "F*** the details don't worry about them" when explaining the consequences of changing something. Must be nice not having to worry about those pesky things.

 

[Train]

If it compiles, ship it.

Thank God for live updates, eh?

Boss: "by the time it hits the client's computer, we'll have an update out!"

Boy has that bit us in the ass.

 

[chusteczka]

My partner always tests his designs when they return from the manufacturer. He will test a design sample after the mold is made and before the manufacturer produces the lot.

While there are some good Chinese manufacturers, it sometimes seems like communicating with children. There are no assumptions; such as quality. Everything has to be communicated and directed from the ground up. The issues are most often due to long distance communication through email compounded by language issues. I will often ask a single question that needs an answer, through email. My Chinese contact sometimes does not answer that question but then asks questions of his own in a different direction, that I cannot answer without the answer to my original question. It gets frustrating sometimes.

Also, the standard Chinese steel Q235 is too soft for most uses, especially fasteners. Nothing beats an American made fastener.

 

[Jeff7]

Chiming in too:
There are many constraints on projects. Sometimes when I'm given a project to design or work on, it can go like this:
"Ok, I should be able to do that. When do you want this, next week, or maybe 10 days?"
"No, I need the drawings out in the shop today so that it can ship out tomorrow."

So that can set the pace for the level of quality that can be built into a design.

Another source of issues can be when it gets to the shop floor. The engineer has some idea of how it might be made, and designs accordingly. The shop workers get the drawings, and decide that they have some slight changes to make, and they make the parts. Engineering never hears about any of these changes, so it's not verified, and it's not documented.

And of course, prototypes cost money. The workarounds are:
- Don't prototype.
- Make a cheap approximation, either of the entire device, or else of just a small part of it.



My own preference is to test the hell out of something to ensure that it works as intended, and if it's something that's going to go into production as it is, I'll ask someone in the shop to make it. That way, I can get a good view of how it will actually be made, and what the result is. After that, I'll see if I can get someone else to have a look at it. I found that double-checking your own work can only be so effective. For example, on a test in my Calc II class, I decided that 3²=6. I double-checked the test, and did not see this error. I was still looking over the test with the same stupid mindset in place that caused the problem initially.
So for that reason, I like to have someone else with a different viewpoint to look at the thing, to help weed out problems.

 

[gophins72]

sometimes it isn't mere cost cutting but careerist management that is responsible. ie. do as much work as quick as you can then ship it, it's the next guy's problem so long as we (my leadership) look great.

 

[Jeff7]

sometimes it isn't mere cost cutting but careerist management that is responsible. ie. do as much work as quick as you can then ship it, it's the next guy's problem so long as we (my leadership) look great.

Or the unfortunate problem of productivity. I created a script that enables rapid modification of a certain model of an existing product at work - instead of recreating a new model for each iteration, I've made it programmable. Now a few simple inputs will cause the model to reconfigure itself in about 20 seconds.
Then people come to expect rapid turnaround when they want a drawing customized. Problem is, I don't have programmable models available or finished for all of our products. Some take 15 minutes to go from nothing to a complete drawing. Others can take a few hours.

 

[RagingBITCH]

It's not their job to test the feasibility nor the user friendliness. That's what QA and consumer feedback is for.