Drawing Check Process 1

[dmg33]

Hi all. I was recently promoted to design lead at a small-midsize company. Prior to my promotion there had been no direct management of this department. Engineering managers made some decisions when it was needed, but never developed any processes or standards for the design department. Now, I'm trying to bring some order to the chaos. We have no PDM/PLM, which makes the task of processing 100+ drawings very painful.

Our undefined, unofficial "process" for a drawing from creation to sign-off is as follows
[ul]
[li]Engineering provides Design a list of drawings/revisions to be created/updated. This list can be provided before or during the drawing creation/update. This list can be modified at any time during the drawing creation/update process.[/li]
[li]Design creates/updates drawings to the best of their abilities and submits printed drawings to the checker.[/li]
[li]Checker checks drawings and returns marked up print to designer[/li]
[li]Designer makes corrections to drawing per mark up and returns corrected print to checker. Repeat until checker is happy.[/li]
[li]Checker gives drawing to engineer to review[/li]
[li]Engineer reviews and marks up drawing and gives to designer[/li]
[li]Designer makes corrections and gives to checker.[/li]
[li]Checker reviews/mark up drawing. If drawing is good, it goes to engineer. Otherwise, it is returned to the designer for corrections.[/li]
[li]When engineer is happy, drawing is submitted to engineering manager. Drawing is corrected until engineering manager is happy.[/li]
[li]Drawing is stamped/released[/li]
[/ul]

To me this seems insanely inefficient. The drawings can just...swirl until all parties are happy and our due dates just sail on by.

Could someone please tell me how this is done in the rest of the world? Or, could you point me to an online resource about this?


Thanks,
D

 

[MintJulep]

In general the process for most of the rest of the world goes like this:

[ul]
[li]Engineering provides Design a list of drawings/revisions to be created/updated. This list can be provided before or during the drawing creation/update. This list can be modified at any time during the drawing creation/update process.[/li]
[li]Design creates/updates drawings to the best of their abilities and submits printed drawings to the checker.[/li]
[li]Checker checks drawings and returns marked up print to designer[/li]
[li]Designer makes corrections to drawing per mark up and returns corrected print to checker. Repeat until checker is happy.[/li]
[li]Checker gives drawing to engineer to review[/li]
[li]Engineer reviews and marks up drawing and gives to designer[/li]
[li]Designer makes corrections and gives to checker.[/li]
[li]Checker reviews/mark up drawing. If drawing is good, it goes to engineer. Otherwise, it is returned to the designer for corrections.[/li]
[li]When engineer is happy, drawing is submitted to engineering manager. Drawing is corrected until engineering manager is happy.[/li]
[li]Drawing is stamped/released[/li]
[/ul]

 

[3DDave]

There should be two paths, one for creating new drawings and one for updating existing drawings.

The new drawing path is always the hard one because people blend drawing preferences with design preferences with design requirements. For example, during the transition to CAD the checkers in my company refused to allow a reference trimetric view on the drawing on the off chance changes would be hand-incorporated and the drafters would be unable to correctly update the trimetric view. Pure drawing preference.

If this sort of thing happens a lot, then start a running list of supposed defects and see if there are some that repeat. There should be some regularity to the notes and the way things are depicted, so that is a basis for a company standard. It then becomes a problem to sort out if the defects are problems with checker preferences or engineering preferences, et al.

Without agreement or the external imposition of specific standards, there is no solution.

For updating drawings, things can be a bit easier - sit the parties down and create a change order document, eliminating the extensive drafting cycle effort from deciding what needs to be done. Then all the checker needs to do is ensure the changed drawing reflects the change order. At that point it can go to engineering to sign and if they don't, then look at why they feel the change order document was inadequate or whey they think what the checker believed doesn't match what the change order document required.

 

[drawoh]

dmg33,

That is a hopeless procedure unless your design work is absolutely routine and repetitive. When you design something, you are juggling multiple requirements. You need to identify and work out the stuff you do not know or understand. This can vary wildly from design to design. You need to communicate with co-workers and end-users to ensure you understand all the requirements. You need to present some sort of document package to the rest of your team so that you can do effective design reviews. A rigid process should not kick in until you are done, including all your drawings and they are ready for checking.

Design checking is a quality control process. You need formal quality standards for drawings to meet, and these must be backed up by management. Your next concern is to find a qualified design checker. If the checker is respected by everyone, his recommendations will be followed.

If it were my call, design checking would be optional...
[ol]
[li]In many design scenarios, design checking will cost more than making mistakes. The design checker can catch mistakes, and provide feedback to management on how designers are performing.[/li]
[li]In an unresolvable conflict between the checker and the designer, you have the option of releasing the drawings without checker approval. Your ass is on the line.[ ][/li]
[/ol]

For QA[ ]purposes, you need an understanding of what the signatures/initials on the drawing actually mean. The checked initials mean that a qualified design checker has determined that the design and documentation meet company standards. I see no need to document the process by which the documents go back and forth between the designer and checker. If they cannot figure this out, they are too dumb to do designing and checking.

3D[ ]CAD is a design tool, not a drafting tool. If you have an engineer scribbling on the back of napkins and a CAD[ ]monkey sitting in front of SolidWorks, you have disabled most of SolidWorks' capability. I do not see a separation between engineering and design.

To sum up, you do not need a procedure. The design process ends when the designer completes the drawing package. The drawings are checked when the checkers satisfies himself that the design and the drawing package meet standards.

--
JHG

 

[ctopher]

Regardless of who checks, there should always be a different set of eyes checking drawings before release to reduce human errors.

Chris, CSWP
SolidWorks '16
ctophers home
SolidWorks Legion

 

[CWB1]

JME but the better organizations I've worked within have always had engineers, checkers, and corporate print quality standards governing the final release. Some folks will say engineers dont have the time or knowledge to create a quality print but IME its vastly quicker than having a draftsman trying to interpret design intent. A half decent engineer should get the print 95% perfect in a similarly quick fashion, if they cant then I dont believe they have the experience to handle design.

 

[3DDave]

The main check I've needed for updates is against inadvertent changes. With parametric models, it is often the case that a low-level change will ripple up in some ignorant way. This is handled by direct comparison via software. Adobe Standard and Pro have comparison built in, but one can also use other tools to create overlays.

The only thing that traditional, good, drafters had over engineers is they spent a long time laying out drawings and making good arrangements of views was a side effect. If they did a poor job, they would have a lot of re-work to do and would either get good or get fired. Now that CAD is prevalent the skill is heavily eroded no matter who does it.

 

[KENAT]

So the basic issue is you have 3 approvals required, Checker, Engineer & Engineering manager and at the moment you bounce back from each step potentially multiple times?
[ul]
[li]What is each check level looking for - the same stuff or different aspects?[/li]
[li]Are they checking to the same consistent standards?[/li]
[li]Do they have the required level of experience/qualification etc?[/li]
[/ul]

For instance, if the Engineer is primarily checking for function and that the designer designed what they wanted designed I'd suggest that step might want to come before the checker. Then the checkers redlines might want to be reviewed by the Engineer for anything they don't agree with before incorporation.

Your issues with swirl are more to do with quality control and/or everyone being on the same page wrt schedule/prioritization.

Using 'schedule' as an excuse for not taking reasonable steps to ensure adequate quality is a bit short sited as it may come back to bite you later in the process with.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​