Hi, Barry,
First thing you have to decide is the purpose of collecting the data. What
do you want the data to show you? If it's only defect analysis data you
want, you will have to have a defect recording system in place.
There is a commonly used form type called a Production Incident Report or
PIR. This form is filled in every time something occurs in production that
doesn't conform to the build plan or right-first-time product build; things
like component breakages, component failures, missing components,
incorrectly assembled components, etc
One section of the form is dedicated to recording the defect code, which is
often subjectively chosen from the available codes by the person filling in
the form, but the defect codes themselves are made up by QA or whoever is
responsible for tracking defect trends. The number of codes can be added
to, but it's inadvisable, as it will interfere with trend analysis to
introduce new codes late on. Make sure you understand in advance the sort
of problems you MIGHT experience i.e. all those that are possible, not just
those that you think you're outfit is likely to encounter.
Codes are are often in two parts - a primary cause and a seconday one. The
primary cause might be a component breakage and a secondary cause might be
because of improper handling procedure, or inadequate packing. Or, primary
might be 'component failure' while the secondary causes might include
inadequate static control, or it was supplied faulty, or it was fitted the
wrong way round, or it was mishandled. I hope you're getting my drift -
make up a list of as many primary causes as you can think of, and then make
up a list of all the secondary causes that could cause or contribute to the
primary cause. A component fitted the wrong way round (primary) could be
caused by an incorrect drawing, or an incorrect planning/work instruction
that was raised from the drawing, or operator error, or placement machine
programme error. A primary cause can have numerous secondary causes. If you
like, you can have tertiary causes as well, but the systems starts becoming
unmanageable then.
Make up codes for all the faults and publish them. Each problem should have
a one or two character code, and the form should be designed so that it
separates the primary code from the secondary. We use a four character
fault code that starts with the project code (could be contract code, PO
code, whatever defines the ultimate deliverable). The next character is the
work area code where either the fault occurred or where the fault was
discovered (Assembly, Test, Coating, etc.), and the final two characters of
the code are the primary and secondary causes.
The codes are fed into a database, along with the assembly information,
serial number, component information, component reference etc., etc. that
absolutely nails down what happened to what precisely, where, when and
how. Then as the information accumulates, you can tell from the codes if
there is a particularly problematic component, or a problem in a particular
work area, or with a particular assembly from teh number of times the codes
come up.
Hope this gives you a start. There are a number of databases on the market
for defect tracking and analysis. In a passed life, we used one called
FRACAS (Fault Recording And Something Something Something), though in its
user-friendliness, its acronym was more descriptive of its useability than
what the acronym stood for. It does work well, though, with patience and
understanding.
Good Luck
Pete Duncan
Barry Gallegos
<barryg@WESTERNELECTR To: [log in to unmask]
ONICS.COM> cc: (bcc: DUNCAN Peter/Asst Prin Engr/ST Aero/ST
Sent by: TechNet Group)
<[log in to unmask]> Subject: [TN] Data Collections
05/04/01 09:21 PM
Please respond to
"TechNet E-Mail
Forum."; Please
respond to Barry
Gallegos
Good Day my Friends.
Help!!!
Am in the process of attempting to put together a data collection and
charting system for my
Electro Mechanical Assembly "EMA" division. I am not sure how to proceed
with this.
defect codes to use, how to collect, where to collect, and how to present.
This is a low volume
environment chassis build, anywhere from 1 chassis to larger runs.
Any help would be so gratefully appreciated. Thank you in advance to any or
all who might
be able to be of assistance.
Barry Gallegos
Process Eng.
Western Electronics
1550 S. Tech Lane
Meridian, Idaho, 83642
P- 208.955-9771
F- 208.955-9755
---------------------------------------------------------------------------------
Technet Mail List provided as a free service by IPC using LISTSERV 1.8d
To unsubscribe, send a message to [log in to unmask] with following text in
the BODY (NOT the subject field): SIGNOFF Technet
To temporarily halt delivery of Technet send the following message: SET Technet NOMAIL
Search previous postings at: www.ipc.org > On-Line Resources & Databases > E-mail Archives
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information, or contact Keach Sasamori at [log in to unmask] or 847-509-9700 ext.5315
---------------------------------------------------------------------------------
|
