I’d like to ask you if anybody of you has ever worked on the reduction of clerical mistakes on drawings. The update of design practice and check list after a root cause analysis doesn’t seem to give satisfactory results.
Just make sure that you’re Measuring the defects from the point of view of your customer. If your customer doesn’t care that there are two periods after the version number instead of one, then maybe it’s not a defect. Talk to your customer (ones that you are close to) and let them know that you’re working to improve the quality of your drawing products. Then ask them how you can improve them.
I am not totally clear about your question, but what brings to my mind is that you would like to reduce the checking time during processing a time.
Basically, a designer gives drawings to a drafts man and the out put drawing is printed and checked by the and highlighted with red( wrong) and corrected there itself and yellow ( correct) , which is right to know the drafts man his mistakes.
Basically a standrad proceedure of checking and coordianting the drawing process and other related engineering to be done by one desinger who conceives the concept.
Hope I am clear. Please let me know is this what you have asked?
How many root causes did you find. I manage a design group of approximately 15 engineers. Typically, there are many root causes to drawing errors. You might first want to seaparate the different “types” of errors into categories, for example: fits, tolerances, GD&T, missing dimensions, wrong specs, spelling, being rushed — not enough time, etc. Then address each one indivdiually. You may find that your not getting satisfactory results because your trying to fix too many different unrelated problems. Drawing errors are really the symtoms of many different problems. Do you have dedicated checkers? How many? Available manpower and their individual knowledge may be a factor in identifying the root cause. Use a Pareto chart to identify (with your customer) to identify the most frequent and most problematic type. Then complete the DMAIC on just that type category. Once you have that type controlled … move onto the next catrgory.
Thanks for your comments.
In my organization the design engineers are often also drafters. They complete the whole design of a machine. The typical problems that we have is just about GD&T, missing dimensions etc. We don’t have people dedicated to check the drawings, ususally the most experienced drafter are involved in the check before the drawings are issued. One of the main root causes is represented by the lack of experience of many young engineers that have to manage very complex drawings of turbomachinery components. This causes a huge cost of failure, especially due to the frequent need of reworks and redesign.
My idea was to put in place a weekly systematic control of all non conformities discovered by manufacturing, ask people to go through a detailed root cause analysis, verify if there’s a lack of robustness and completeness in design practices and check list, and in the end ask them to propose corrective actions. Monthly I was thinking to organize a meeting with all people to share the lessons learned.
The general idea is to share with everybody, especially with people with no experience, all lessons learned, in order to aware them about the typical problems they can encounter, absolutely avoiding to blame them.
Do you have any experience on this?
I came form a semiconductor equipment that also designed very complex and high precision machinery. I had approximately 22 mechanical engineers, electrical engineers, and software engineers. The engineers did most of their own work — layout, models, details, checking, etc. Essentially designers make poor checkers and detailers. I would still separate different symtoms into categories — missing dimensions could have a different “root cause” than an incorrectly called out datums structure.
My expereince with missing dimensions is twofold: (1) that the designer or checker only checks what they “see”, not what they “don’t see,” and (2) they just don’t use a structured methodology specific to looking for missing dimensions.
One suggestion that has worked for me is to have force them into a structured and “disciplined” methodology. You will probably find resistance to this because it will probably take longer, but quality will improve.
This should be the very first task in checking. The checker (the designer in your case) should be checking using colored pencil on a print of the drawing. Igonre checking anything other than “are all surfaces dimensioned” — don’t worry about whether the dimension is correct at this time. Color the surface (not the dimension) ”yellow” if there is a dimension for it — be very critical at this process. If a surface does not have a dimension, draw a set of dimension lines without a dimension in “red” pencil. Fill in the missing dimension later.
The same follows with fits and clearances, datum structures, true position of holes, etc.
If a problem exists with general GD&T knowledge, you may want to look at self-paced training material through SME; and devote a certain amount of time each week in a group discussion on only one subject. People learn and retain much more information in a group environment.
One of the root causes of the drawing problem is that your engineers just don’t take it seriously. Usually I have found some engineers think of detailing and checking as necessary evils — get it done as quick as possible. Try using the metrics from the errors as a feedback and performance issue if necessary.
The forum ‘General’ is closed to new topics and replies.