Reducing Mistakes in Designs 5

[natew2006]

Hello all. I just started my job (3 months ago) as a design engineer at a small company. Since I am new to the company, I don't fully understand all the details in manufacturing and component selection. Therefore, I sometimes make mistakes in and/or leave notes out of the production drawings. I have some reference materials to work by, but it seems like we mostly work on precendent. My supervisor gets frustrated that I make these mistakes, and it is starting to worry me about their perception of my ability. Does anyone with more experience have any suggestions or advice on how I can prevent these mistakes? IT seems impossible for me to completely eliminate the errors with so many details.

Thanks in advance

 

[MikeHalloran]

1. Ignorance is correctable.
2. If you made the same mistake twice, your supervisor would be justified. Making it once is the price for hiring someone inexperienced in the first place. The last guy who had your job could probably do it without a lot of mistakes just before he got fired or quit in frustration.
3. If you don't understand the manufacturing process, get out in the shop and watch it happen.
4. Take notes. These are for your own benefit. Write down who said what while it's still fresh in your mind. Don't bring a clipboard out in the shop for this purpose. Ask questions, listen, go back to your desk and record the answers.
5. Reverse engineeer your own products, to find out why they are the way they are. The old timers in the shop can tell you in what ways they used to be better, but they won't really understand why. That's why you went to engineering school.




Mike Halloran
Pembroke Pines, FL, USA

 

[Pressed]

The basic catch for making and checking your own work:

Make a copy print (normally letter size) and go through each small detail to check it's correct and cross it out with a highlighter then go to the next small detail and so on. (notes, dimensions, title blocks, etc...everything or everything in question)

To reduce the errors of mind-set, do the check the next day.

Other than that get relevant books on your subject and read/keep em handy.

==========================================
Business Card

http://mech.e.tripod.com

__________________________________________
Cycle Heaven.......

http://www.tailofthedragon.com

 

[calguy07]

hey, i'm in a similar situation but i'm 6 months in (electrical design). the key really is to learn from you FIRST design. look at the mistakes you made and just be aware of them in subsequent designs. each one will get better. also be sure to look at other designs done in your company on similar projects. no two projects are the same exactly but looking at examples certainly helps..but i guess you knew this part already. make sure to always ask questions. not sure its right for the supervisor to get frustrated after only 3 months whether you make the same mistake twice or not (as long as its not EXACTLY the same mistake though). no one's perfect.

as far as checking goes, i agree with the highlighter method. it definitely helps to see the actual print as opposed to staring at the CAD program file on computer. also, its not easy, but try not to get too frustrated. i've found that no matter how much i check my drawings, i always miss something. that's why we have someone else checking our designs.

 

[tygerdawg]

It's idiotic and a sign of poor management to expect someone with 3 months experience to be perfect, then get frustrated when they are not. (The same goes for someone with 25 years of experience, only less so.)

Mr. Halloran has some excellent suggestions, but self-learning can only go so far. Admit your ignorance and ask to be partnered with a more senior person. That's the "mentoring" process. Your boss should have already known to do this since engineering design is a creative process to be learned, practiced, and polished to a high level of compentency. If he does not, then shame on him.

TygerDawg

 

[MadMango]

Yes, checking your own work will only get you so far... not seeing the forest for the trees and all that. Get someone else to check your work if someone else is available.

If you have teh luxury, look at existing drawings/designs that are similar to what you have created, and do comparative checks to see if there is something obvious that you have omitted.

I know you are probably worried since it is 3 months, ie 90-day probation. If there is no training program in place, it is a short comming of your employer, not you.

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of Eng-Tips Forums?

 

[ScottyUK]

Expanding on Mike's advice, make sure you visit the maufacturing line when your product is being built. Ask the people who are building it to tell you where the problems are, what the limitations of the machines and tools they have are and ask what would make things better. Ideally get them to show you so you see the problem yourself instead of hearing about it.

See if your boss will cut you some slack to spend a week or two on the shop floor building the things you design. When I worked in hybrid circuit design I had spent time on the line beforehand, so I knew most of the limitations of the processes and machines and deliberately designed things in a way that made things easier for the shop floor. It was nice to get the occasional 'thankyou' and the rest of the time the lack of flak from the line engineers was equally rewarding.


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[lz5pl]

natew2006, in my first working day 24 years ago my Boss said: Our job (commissioning of high voltage substations) is 70% routine and only 30% science. Hence look for advice not only from your mentor, but also from experienced technicians.
I think that this advice is applicable to most engineering fields, so I will join previous colleagues - the best way to learn is practice, where your product is manufactured. Don't be ashamed to ask for opinion guys on production line.

Work and check your designs as carefully as possible, but don't expect ideal results. I work my profession 24 years and still I sometimes find mistakes in my own design when I go to the site for commissioning. This makes me furious on myself, but that's the life. I remember an old story about ancient Japanese artists: When they finish some nice figure they breake a small piece of it - because only Gods could create somethink ideal!

And the last: never stop to read and study, at least until you enjoy your job (when you stop enjoy it is time to leave it). Study from your mistakes and from other's mistakes. It is not a shame to make mistake, it is not good, but it is possible to make the same mistake for a second time, it is not acceptable to make it for third time!

------------------------
It may be like this in theory and practice, but in real life it is completely different.
The favourite sentence of my army sergeant

 

[cadengnr]

I could have really used this thread a couple of days earlier.(See "Reality Check," if you dare.) True, I can't stand my boss, but what I can't stand even more is having anyone on the floor, engineer or machinist, catch a mistake when I felt I should have known better.

Some thoughts:

1) Make a checklist of things to help you go over your drawings - general notes, sheet metal notes, all the balloons are there, and - at least make this mental note - everything has an overall width-height-depth. Sometimes we get so into the trees that we miss the forest.

2) Don't feel too much pressure to get it off your desk. If you can, once you're "finished," put the job down and start something else before going back for a last pass. If the weekend's upon you, do your final check Monday.

3) Yes, there's something about good ol' paper. Print it out and redline it. Especially if it's big-size, print out a scaled A or B-size so you'll have an easy bird's-eye scan before taking up plotter time and ink.

4) Absolutely - seek help from the "floor": the machinists, etc. will respect you more for it (and you in turn will find out how much more out there you really don't know!). Communication with them is crucial. As for the boss; many bosses rather wouldn't think about learning curves and personnel issues; they just want the job done. If you think he's not going to allow you the time, you might want to bring the topic up with him, and ask who on the floor are the best people to consult with. That way, if he deserves the position he's in, you'll have done something to appear self-aware, pro-active, and positive.

5) With experience, you'll notice more mistakes in other people's drawings. I'm talking stuff from General Dynamics, Lockheed, or Boeing that's on Rev Q or later. Why does this happen? Few people take as good a look as the first guy, for one thing. In some cases, like for a cast part, the cast/mold makers take the mistakes out of the tooling itself, and the drawing error stays while good parts get churned out for years.

6) Yes, mistakes can cost dearly, especially on a carrier deck. But even veteran pilots bolter or fail to catch the third wire. 'Doesn't mean they never stop aiming for it or quit flying, Mav'.

If "Top Gun" ever had a moral, this might be it: Take responsibility for your mistakes, learn from them, and don't stop trying to improve. But don't spend too long beating yourself up about them. Before you think you're no longer cut out for something, get back in there and keep flying.

 

[GRoStruc]

A lot of good advice on here already.

Agree with the yellow marker/red pen on all drawings check to ensure that the drawings are exactly as you've intended, but remember, a correct drawing is not a correct part/assembly etc.

Get out into the machine/fab shop and find out how things are made / built. Aside from ensuring you get the details right on future designs take note of how things are REALLY done and tailor future designs with this in mind.

Small design changes can make a big difference in cost and time further down the line and you’ll be amazed at the things people don’t question (i.e. dumb mistakes that are dutifully carried out by people that should know better).

Question details on drawings when you can rather than blindly go on precedent. You’ll gain a much better understanding of why and what notes/details go on drawings, and probably put some of the older guys to shame, remember, knowledge is power!

If you get the chance as well get any feed back from the end users. Great design and ease of manufacture is one thing, but if it doesn’t do the job properly then it’s pointless.

Remember as well, we’ve all been 3 months into a new job made mistakes (still do at times!). You shouldn’t be expected to be perfect straight off, ask for plenty of advice and learn from your mistakes

 

[HDS]

Does you company have standards for you to follow? For example is there a template for a sheet metal drawings with notes ect that are a starting point? Is there a document that lists the tooling available and the preferred parts to be used?

If not consider making that your next project. It will help you learn and help everyone else be more efficient.

At the very least make additions to the reference materials where they are needed?

 

[csd72]

HDS basically stated my point,

When you get a new job, ask if there is a similar project that you can use as guidance.

This is not cheating this is just a common sense reference to how it has been done in the past (it would be hard to design automobiles if you had to reinvent the wheel every time).

Of course you should question every detail on the old drawings but it gives you a guide.

Also if there is a lot of repetition in what you do, write yourself a check list to jog the memory and make sure you dont miss anything obvious.

Dont let them get to you, just do your best, everyone had to start somewhere.

 

[KENAT]

Checking is probably the most significant thing that can help you. I’m going to assume you’re in the world where you create 2D drawings and don’t primarily rely on MBD.

Self checking is a good start. As others have said print out a copy of your drawing and go over it, make a check list if it helps but generally look at things like are all required views and dimensions to make it present, all notes required present, does it comply with whatever drawing standards your company has… I’d use colored pencil not marker, that way you can erase it if need be. Use different colors to indicate things that need deleting/changing (typically red), things that stay the same (yellow) and then comments or questions for you to come back to later (blue or green). In places that are strict about it Designers should normally only use green with red yellow & blue reserved for Checkers but this doesn’t sound like your place.

Repeat until you can’t find any problems.

There are two obvious problems with self checking:

1. You can’t really check what you don’t know/understand/aren’t familiar with etc

2. Being intimately involved with the item you can’t always ‘see the wood for the trees’ as others have mentioned. Leaving it on your desk overnight and looking at it the next day helps but even then you’ll still miss stuff from being overly familiar with it. If nothing else on notes you may well understand what you mean but perhaps no one else will!

The first can be corrected by learning/training/experience but these take time (others have made some good suggestions on how to improve this area so I wont say much however reading relevant books, spending time with more experienced staff and/or on shop floor and familiarizing yourself with any relevant industry/drawing standards are good ideas)

This is why any self respecting (in my opinion) Engineering organization has some kind of checking process by at least one other party.

Ideally this will be dedicated checker(s) with many years of experience, familiarity with the relevant drawing standards etc. (This type of checker is a dying breed though as managers seem to think that with CAD etc we shouldn’t make mistakes and so independent checking isn’t needed any more.) 2nd best is just to have the most qualified staff member available do the check. 3rd best is just to get any warm body that is vaguely familiar with drawings to look at it.

This actually achieves two things, first it catches the majority of problems with the drawing in hand. Second it provides a source of information to learn from. Don’t just blindly incorporated the check marks, try to understand them and apply them next time round, any that you don’t fully understand ask about and if necessary research. I virtually learnt to draw by having my initial drawings bled (bled - due to the amount of red ink/pencil looking like blood) over by a couple of the senior design Engineers. I early on set it as my goal to get a drawing through without getting any red on it. To this day that remains a rare feet!

A word of warning on copying previous designs. If they were good designs and technology hasn’t moved on much etc then this can be really useful. However, blindly copying designs, especially if you don’t fully understand them, means you’ll also copy any errors, problems, insufficiencies etc and perhaps not incorporate changes that might be beneficial.

Also a word of warning about relying too heavily on what the shop floor tells you. They primarily care about ease of manufacture. They typically care less about functionality or inspection/quality control. As well as talking to the shop floor talk to quality &/or whoever does inspection. They also for the most part don’t have the level of technical/academic engineering knowledge that graduates (or even people who’ve gone through an apprenticeship or similar) have. This isn’t an excuse for arrogance but a warning to consider carefully what they tell you, not just blindly do it.

Not meaning to down play DFMA but when creating a drawing the most important thing is probably ensuring performance/functionality of the component. Second is ensuring that it is possible to verify performance/functionality from the drawing (i.e. supports inspection). Creating the drawing so that the part is easy to manufacture/ drawing is easy to manufacture from comes after these 2, you have to make sure the part can be made and try to take into account DFMA principles but this comes after basic function and verification of function.

An obvious example is in dimension schemes, dimension schemes should be created to aid inspection more than manufacturing. A lot of people dimension the way they thing the machinist will want the information, to centerlines or centers of arcs or other theoretical points without considering how these dimensions are to be checked, especially after the piece is taken off the tool. Inspection dimensions can always (at least I can’t think of any exceptions) be converted into the dimensions the machinist needs however it doesn’t always work the other way around.

Just my 2C worth (or looking how long it is maybe 10C)

 

[ewh]

Excellent advice, KENAT!