Drawing Checker/Approver Qualifications 8

[KENAT]

Sorry for yet another post but:

For those of you who have a formal drawing check process (not just dedicated checkers) what are typical minimum qualifications/job requirements.

Just to narrow the field, I don’t just mean a vague peer review but thorough checking for completeness, correct GD & T, standards compliance, basic form/fit/function, tolerance analysis, DFMA etc.

I tried looking at job postings but almost all of them listed previous checker experience as being a requirement and I’m just looking for bare minimum.

It’s come up because we only have one ‘approved’ checker, me (and frankly my qualifications for the position are questionable) and people are complaining about the back log and asking why they can’t check etc. We’ve had at least one Engineer send several messages saying that someone with a Masters prepared the drawings she with Bachelors already reviewed them, why isn’t that sufficient. I just finished checking the first pack from her earlier today and, lets just say they weren’t good.

We have a check policy that says my boss maintains a list of approved checkers, I’m the only one on it for now. We want to formalize our requirements for being a checker so that we can defend our stance of not letting just anyone with a pulse check and also hopefully so it can help us find someone else to share the load.

For starters I’m thinking:

1. Minimum 5 years preparing drawings to ASME Y14.100 (or equivalent).

2. Skilled in the application and understanding of GD&T (ASME Y14.5M-1994), preferably at least GDTP Technologist level or broadly equivalent combination of training and experience.

3. Experience with ‘Worst Case’ stack up tolerance analysis including impact of GD&T.

4. Familiarity with common manufacturing processes preferably with knowledge of DFMA principles.

5. Good communication skills to explain drawing changes, standards requirements and represent documentation requirements at meetings etc

6. Internal candidates should have a proven track record of complying with relevant company policies and procedures.

I have my doubts about academic qualifications (other than GDTP etc) having much direct relevance though maybe a minimum of high school wouldn't be a bad idea.

Any suggested changes or additions?




KENAT, probably the least qualified checker you'll ever meet...

 

[ctopher]

I think you have a good start.
For #1, I would suggest min 10 years.
#5, good communication (written AND verbal skills)
I think I would also add a good knowledge of machining practices.

Chris
SolidWorks 07 4.0/PDMWorks 07
AutoCAD 06
ctopher's home (updated 04-21-07)

 

[KENAT]

Thanks Ctopher, I was originally thinking 10 years but don't want to disqualify my self, I only have approaching 7 years.

Emphasis on machining is good idea, as that's how most of our parts are made. Also written & verbal, good point.

I'm pretty sure that whatever I come up with either I'll be disqualified or, I'll have to water down requirements so that I'm not.

KENAT, probably the least qualified checker you'll ever meet...

 

[drawoh]

KENAT,

I cannot see how someone can check mechanical drawings without extensive experience at mechanical design. This gives them the training at GD&T and other drafting standards, and DFMA. Extensive design experince in-house means they are familiar with your particular design problems, and they have a track record of good judgement and getting things done. You can ask the fabricators how good their drawings are.

I am trying to think back on how competent and knowledgeable I was with five years experience. I agree with ctopher about the ten years.

Your tag line implies you were saddled with this. You could state that five years is the minimum acceptable, but ten years is strongly preferred.

JHG

 

[prohammy]

Ken,

Maybe number 7 on your list should read something like, requires the patience of Job (biblical character) and the hide of an elephant. Coming into a company as a design checker requires both (as I am sure you know) and you need to be sure that the person you get can stand up to both the natural resistance/obstinance of the designers (I freely admit that I have been like that in tha past but I have seen the light) and senior managers (usually outside an eng. dept) who want things 'now not tomorrow'

Good luck finding the right person

Kevin

“It is a mathematical fact that fifty percent of all doctors graduate in the bottom half of their class." ~Author Unknown

"If two wrongs don't make a right, try three." ~Author Unknown

 

[ewh]

I agree that familiarity with machining is desirable, as is extensive design experience. You would also want someone who knows how to produce a GOOD drawing, and is well versed in GD&T (inspection experience perhaps?). You need someone who can manipulate the CAD files to ensure proper modeling techniques are followed.
If you add up all of the desired characteristics of a good checker, you will severely limit the number of applicants available.
This is probably the reason why the best checkers that I have known were not young. It takes time to gain quality experience. You might have to rank your requirements as to what is most important, and (I hate this) settle for the best you can get, and give the winning applicant the support necessary to improve to the level that you desire.

 

[KENAT]

Thanks gents.

I do not consider myself qualified/experienced enough to hold the position but as I've mentioned before our Senior Design Checker got laid off and I've been determined the least bad candidate to fill the void.

Drawoh, agree about the experience. Several people here have stated that we should be able to train checkers which I dont believe to be true. We're trying to formalize what the bare minimum requirement is for experience/qualification. We've only recently introduced industry drawing standards here and outside our department most people don't really follow them so hence the proviso on internal candidates and the fact I emphasize 14.100 in the experience.

Prohammy, you're not wrong. We actually have that (phrased differently) in our job description for hiring internally. Sadly given we just lost our checker to lay off this isn’t actually for a job posting for an external hire.

The purpose of this description is so we can end up with an internal ‘job description/minimum qualification’ to defend against the idea of any old Joe doing it. I plan on saying it was prepared after review typical industry requirements and consulting with industry experts (that’s you guys)

The other idea that’s been mooted is that we approve some existing Engineers/Designers as lower level checkers for simpler tasks.


KENAT, probably the least qualified checker you'll ever meet...

 

[KENAT]

Didn't see your post ewh, your points are well taken especially re 3D CAD. We aren't formally checking models yet but we have identified it as something we'd like to do and on one major program it's getting done to some extent.

As regards the settling, which I too hate, I haven't seen anyone here outside my department who comes close to being marginally acceptable!

I don't come anywhere near the previous incumbant of this position in terms of qualification or experience, I knew this before grudgingly accepting it and get reminded of it daily. I'm not yet 30, he was past retirement age. There's no comparison.

But I'm trying...

KENAT, probably the least qualified checker you'll ever meet...

 

[KENAT]

differently) in our job description for hiring internally.

Doh, what was that about strong written comunication skills.

Should have been "... hiring externally..."

KENAT, probably the least qualified checker you'll ever meet...

 

[drawoh]

KENAT,

You cannot teach ten years of design experience.

Having said that, there is a training issue with design checkers. You need to be very organized at handling information. With a good, methodical process, you can check a drawing package a day. Without it, you could take weeks.

If the checker has design experience, they will be familiar with design tools. If a drawing package is well executed, it is extremely unlikely the designer does not understand CAD thoroughly. If the package is a mess, it is likely the designer is incompetent with CAD, but sorting this out is not necessarily the checker's responsibility.

The most important thing you need is the backing of management.

JHG

 

[ewh]

drawoh,
It seems that you don't place much value on a well constructed model. With the power available in todays CAD systems, this can be a serious oversight. Properly constructed and maintained, you can make a change to one part and have all related parts update automatically. These changes can be simply changing an expression.
If the models aren't properly constructed, this simple change may turn into a complicated change involving MANY more man-hours to implement.
I agree about the importance of design experience. I just wanted to emphasize that, while a model may "look" good, that it pays off in the long run if the model is indeed "good".
There are very many CAD jockeys out there that can create models and drawings in very short time, but not all of them are methodical enough to create robust models (similar to the lack of good drawings we see today). If your product involves any complicated geometry or assemblies, it is better to establish modeling criteria from the beginning.

 

[drawoh]

ewh,

Put yourself in the shoes of a good design checker who is admiring a clear, accurate, standards compliant drawing package. The CAD model is a link in a chain that, apparently, is working.

It is possible that a CAD operator could produce a throughly screwed up CAD model, and that an experienced and very skilled drafter could turn the mess into high quality drawings. This would be managed pretty much the way things were done back in the drafting board days with a calculator and a scratch pad to make sure all the dimensions lined up. This would be a pretty deranged waste of company resources. It would be a lot easier to fire the CAD operator.

High quality drawings are produced by well organized, methodical, knowledgeable designers. High quality CAD models are produced by... ?

When I finalize my SolidWorks fabrication drawings, I go into the assemblies and I delete all the external parametric constraints. I do not want features of a finalized fabrication drawing controlled outside the CAD model attached to the drawing. SolidWorks updates read-only CAD models in RAM. The dimensions on your fabrication drawing can vary depending on what other files you have open when you print it.

In the real world, you are not allowed to change form, fit and function of your fabricated parts. You certainly are not allowed to do it randomly and without warning.

JHG

 

[ewh]

(getting on soapbox)

"Put yourself in the shoes of a good design checker who is admiring a clear, accurate, standards compliant drawing package. The CAD model is a link in a chain that, apparently, is working."

The key word here is "apparently". If you only need this package for a design that is mature and will need no further development, then I agree with your point. If there is any possibility of this package being the basis of a new design, then the model is a very important link, and deserves a more thorough examination than just having a clear, accurate, standards compliant drawing. It also effects CMM and CAM.

"High quality CAD models are produced by..." well organized, methodical, knowledgeable designers.

"SolidWorks updates read-only CAD models in RAM."

This sounds like a software issue, and if proper controls are in place, this is not an issue with NX. I am not familiar enough with SolidWorks to comment further.

"In the real world, you are not allowed to change form, fit and function of your fabricated parts. You certainly are not allowed to do it randomly and without warning."

In the real development world, you are EXPECTED to change form, fit and function as the design progresses. You have to have a record of these changes, and have a robust PLM system in place to document these changes. My point is that, especially in development, it is much more efficient to change a detail and have a domino effect of updated parts result than it is to change each and every part individually. These changes would all be recorded by default, and you would be aware of what parts you are changing. You must also have good communication within the company.
An aircraft is a highly complex mechanism, and getting it approved and flying in a timely manner will make or break a company.

Anyway, this is getting off topic, and different industries do have different needs.
(getting off of soapbox)

 

[fcsuper]

As humble as you are, KENET, I would consider you the most qualified person I see regularly posting here.

Matt
CAD Engineer/ECN Analyst
Silicon Valley, CA
sw.fcsuper.com
Co-moderator of Solidworks Yahoo! Group

 

[KENAT]

We typically have bad models and drawings.

Sometimes a good model and bad drawing but this is very rare.

What will happen is that someone who knows what they are doing with drafting will update an old drawing to look fairly good, but leave the model in a state. Low and behold when you want to make a development based on that part (or when it doesn't effect form fit function etc a change to that part) you hit big problems.

We have got some basic modeling rules/guidelines [such as the model dimensions/dimension scheme being the same as the drawing and the thing about deleting parametric links prior to release] but few follow them and in our priority list correcting this has had to take a secondary place to getting the actual drawings to an acceptable standard.

I do not assume that if the drawing is good the model is good. If nothing else the person preparing (or at least finishing) the drawing isn't always the person that initially created the model. Drawings are laid out for you to see any issues, with models problems may be hidden although our CAD has tools for finding some of them. However I don't always have the time/remit to do much about it.

As to the off topic issue about the benefits/problems of parametric links:

During initial design/development phase, before hardware, incredibly useful. During Prototype stage when changes are being made still has major benefits but can start to cause problems. During production/sustainment phase, problematic (this is what I took drawoh to be talking about).

I don't so much mean in terms of creating assemblies out of parts (though occasionally this can have issues) but in terms of features in one part being driven by links to features in another.

You change a feature in one part and it changes another part that isn't immediately obvious. This is especially problematic if the part you're working on is used on a number of unrelated assemblies. Especially if one half of the 'pair' is used independently from the other in some applications.

Of course this is minimized by following proper revision protocols but problematic at times none the less.

drawoh - easier to fire the CAD operator, in my dreams. Plus then there'd be no one left except me and one or two others. Backing of management, I almost fell out of my chair laughing on that one, with one or two exceptions (my immediate boss is good as he was brought into the company to improve documentation) management (and other engineers) see the checking process as an unnecessary inconvenience/delay.

fcsuper, It’s easy to be humble when the others around you are (mostly) more experienced and knowledgeable. Of course others may have different opinions as to my humility, you only see my virtual self.

ewh, I agree about CAD jockeys creating models/drawings fast but when you come to do anything with them later, night mare. I lay part of the blame on typical CAD training. It’s usually all about what the software can do, not what you should do with it. Perhaps it’s a subtle distinction but an important one.


KENAT, probably the least qualified checker you'll ever meet...

 

[121560]

As a design checker I agree with a majority of whats been conveyed above, "the checker does not need to be the sphincter at the end of the flow chart". Automate by creating a begin environment, for example create master models in solidworks for various product configs, identify
internal(you) and external(customer) constraints, model and create drawings (templates). For example we, the company I work for we identified approx 60% internal constraints ( we control) and 40% external constraints(customer control) standardized models and drawings where by designers and checker followed est guide lines. Basically check time reduced by 60%. Of note the drawing config templates also incorp standard notes for specific manufacturing process, finish,material and id process specification (hyperlinked) for ref when creating drawing from templates. Also all this was collaborative design and project engineers, drafters ect.. with periodic design reviews of cad std's.
Also created automated excel BOM spread sht linked to master models. Bottom line start with one product master model and cookie cut to other similar product configs.

 

[ewh]

That is a great way to start contolling your documentation, 12156!

 

[KENAT]

Thanks for the ideas 121560.

We don't use Solid Works, though I believe I understand what you are talking about.

We do not really have semi standard parts which can be set up as templates. However, we are applying some of these ideas to the top level tools assemblies where possible.

Cables is one area where we might be able to set up a more detailed template (at least drawing), an old one exists but needs updating to reflect drawings standards.

We already have the standard notes, it's not very elegant (inserted word file) but is workable, some need values adding and typically about 3/4 end up getting deleted. However, a lot of people will change it however they feel like or even delete the whole block.

When we first introduced the CAD standards we had get together with representatives from each department to get agreement. Unfortunately a lot of them don't see standards as being of value or play nicely in the meetings nodding their heads etc and then completely ignore it in practice.

Master BOM & configuration is done by our MRP system.

So long as the majority of Designers and their immediate supervisors/project leaders don’t seem to care about the standards, or in some cases even about creating high quality documentation, Check does end up being a bottle neck. Hopefully things will improve as more people are getting their drawings checked and start to learn how drawings should look but I’m not optimistic of it being quick or easy.


KENAT, probably the least qualified checker you'll ever meet...

 

[JsTyLz]

Word of advice I found out a while back...
"Don't waste your time making the process idiotproof. There will always be a better idiot."

 

[Standing]

Checking is a thankless job and not many want the job. It is not easy telling someone they made a mistake. Engineers do not want to check drawings for standards and etc. So they do not make good checkers. I would rather hire someone who wants the job, because they enjoy it. Rather than hire someone who is experienced and hungry and hates the job.

I agree with the requirements for checkers. That would be a nice, real world, place to work. The issue I have is that a checker can just sign it off. If something is missed, “I’m sorry I missed it”. In the 80’s I was getting $18 per hour as an engineering checker, not drafting checker before CAD. The checker I worked with was experienced and getting $28 per hour. His way of proving himself was to have drafters move dimensions around the drawing. We do not have checkers here were I work, non-military and non-medical. CAD has solved most of our issues.
I had an engineer check one of my (4 E size sheets) design changes to a pipe bender. It took him 30 seconds to sign his name as checked. You get the idea.


Bradley
SolidWorks Premim 2007 x64 SP4.0
PDM Works, Dell XPS Intel(R) Pentium(R) D CPU
3.00 GHz, 5 GB RAM, Virtual memory 12577 MB, nVidia 3400