Re: Overlapping Operations days -- BPCS-L

Santa

In this post & earlier ones, I have identified several files that you might
want to study the contents of ... what is this or that field used for in your
environment.  When studying a file that I am unfamilar with, I like to do a 3
step process.
1. DSPFD FLT F4 *PRINT whatever the file is ... IBM layout reference
2. RUNQRY FLT F4 *YES record selection on bottom
3. Another side session simulataneously WRKQRY looking at the file layout

WRKQRY supplements DSPFD for rapid navigation to make sure I am on the right
field name ... typically in RUNQRY I am looking at some column of data "what
the heck is this" I ask myself, about the population I am seeing ... I use
RUNQRY F19 F20 sideways to count ... this is field # 25 I am looking at ... I
use WRKQRY F19 vertical to count ... field # 25 is called _____ whatever & I
page through the DSPFD print out to jot down some notes there.

I speculate about the relatationship of field # 25 data to various other
fields & the reason for RUNQRY selection *YES is F12 back to selection screen
multiple times to do selection criteria to check out various patterns of data
population that I am seeing.

When you gave your extreme example, were you looking at ALL THE DATES in ALL
THE FILES ... FSO FOD FMA ... in my experience the dates are NOT identical
but are backward scheduled on multiple operations.  FSO says when the shop
order needs to be completed, but FOD has start dates that are scattered over
a period of several days, depending on the routings spelling out the rate at
which the work can be done & the volume that needs to be done.
FSO has one record per shop order
FOD has one type-1 record per operation
FMA has one record per material consumed by the shop order

Remember that we are not using negative days in our company, but we are in
reality.

What Time Basis Code are you using?
Do you understand Time Basis Codes in the routings?

For example, one says that each unit can be made in 0.0001 time period,
another says that x,xxxx units can be made in 1 time period.

Are your work centers setup for 8 hour days, or have you made allowances for
people getting potty breaks & other reasons why when we pay people for an 8
hour day they really are working a little less than that.

Your corporate goals may not be quite the same as ours, and my role in this
also a bit different.  Our goals are to make a profit & deliver product to
customer when we promised delivery & to have zero tolerance for quality
problems in the delivery.  Now this means that there is great interest in
such topics as cost variance ... we compare price / standard cost / actual
cost.  My role is to make the process easier on the people working with the
system.

In our labor reporting we seek to accurately capture the hours it took to
make the product, the material consumed, the scrap, the down time.  We have
also modified the process to reduce the work load needed to enter labor
tickets, to reduce the risk of human error, and to add capture of some
information to other data base elements ... mainly we track tools neccessary
to the production process, such as molds, then have later reports showing
scrap rates by tool-id which tells factory management which of those tools in
greatest need of preventative maintenance.

Without modifying CST900 itself, I have made modifications to the tracking
process.

Before CST900 runs, I have a query recreate a summary query/400 work file
that contains some FSO information on the orders coded for purge & what the
actual cost is on those items before the purge.  Then after the purge, I run
reports on the new actual cost variances on those items, and by how much the
actual cost has changed thanks to the CST900 run.  This is all in a jobstream
CL in which one is before CST900 & another after CST900 in which we
periodically adding new clue reports to the collection.

Also before CST900 we run a program which is a clone of CST270 using a
different logical over FOD to eliminate the garbage from the report, that
SSAX says is supposed to be there (treating type-3 lines as if they were
type-1).

The queries tell us which items in recently closed production had bad cost
variances.  We run a search on CST270 clone to see the whole WIP cost story
on those items ... invariably 80% or more of the variance is in one
subcomponent.  We search CST270 clone for that subcomponent.  We keep CST270
clones for several past CST900 runs.

90% of our cost variances, analysed by this means, can be traced to labor
reporting in which the actual rates on some subcomponent just came nowhere
close to standard.

If you want to take a look at this kind of information, without doing all the
gyrations that we have indulged in so far ...

FLT has labor history ... it has lots of information about RATES & very
little about costs ... how much work was done in what time period & how does
that compare to the standard ... the cost in our case comes from the CEM file
on the clock # used in the labor reporting

ITH has actual cost & standard cost of the inventory transaction ... we use 3
kinds ... receipts of completed materials, issues of consumed materials,
consumption of materials when rejects.

There is a labor ticket # assigned by JIT600 & that goes into both the FLT
transaction & corresponding ITH transaction on same reporting, so you can
link the two by the ticket # & get the whole story on the rates & costs
associated as of the time the labor reporting went into the system, and you
can also get the date this was actually reported to the system.

The date is important because when CST900 updates costs, it uses the latest
costs of the material being consumed, including costs just updated by the
previous shop order purged.  Thus of interest is to do a report on a
particular item's actual costs by date time processed from labor thru ITH to
show how the actual cost jumps periodically due to CST900 impact.

MacWheel99@aol.com (Alister Wm Macintyre) (Al Mac)
BPCS 405 CD Manager / Programmer @ Global Wire Technologies Incorporated
http://www.globalwiretechnologies.com = new name same quality wire
engineering company: fax # 812-424-6838