Good afternoon,
Oliver Wight invented a phantom item solution to your issue.
It works operationally but requires "wide-awake" attention to the
cost accounting effect. To discuss this let's create an example
as follows:
Part A is currently a sub-component to part ENDITEM.
Part R will replace part A in the future.
The replacement of part R for part A is universal and permanent.
Your replacement date is not forced by a required change (e.g.it's not
a change to correct a product safety problem nor a change to comply
with a dated consumer promotion printed on the container nor
any other such "can't move the effective date" constraint.)
Then try this in your test environment:
++ Insert new part R in the BOM as a child under existing part A.
++ Set up a reasonable purchasing lead-time for part R.
++ Change part A into a phantom item (zero lead-time, zero
safety stock, lot size = 1, can't buy it, etc.). It becomes what
Oliver Wight called a "pass-through" item.
The expected test results are:
When the MRP math shows inventory for part A running out, MRP will
no longer generate a planned PO for part A, but it will throw off planned
POs for part R that are consistent with the purchasing lead-time for part R.
If orders for ENDITEM run ahead of plan, then the next time you run MRP,
it will advance the planned release date for the first part R purchase
order. Conversely, if orders for ENDITEM decline vs. plan, the next
MRP run will push off the suggested release date for the first part R
purchase order.
Three database janitorial responsibilities:
1. When all of the inventory for part A is gone for sure, remove it from
the BOM and promote part R to be a child right under ENDITEM.
2. Once part A is out of the BOM, you probably want to de-activate/
soft-delete part A from the item master file to prevent any prospective
accidents. For a discussion on that supremely frustrating chore, go
here: www.unbeatenpathintl.com/iu/source/1.html
3. The cost accounting attention is needed to make sure a roll-up
doesn't include the cost of both part A and part R.
Warm regards,
Milt Habeck
Unbeaten Path
www.unbeatenpathintl.com/2.3bw/source/1.html
North America toll free: (888) 874-8008
International: (262) 681-3151
mhabeck@xxxxxxxxxx
+++++++ +++++++ +++++++ +++++++ +++++++ +++++++
From: Bailey, Dick
To: bpcs-l@xxxxxxxxxxxx
Sent: Sunday, April 15, 2007 1:58 PM
Subject: [BPCS-L] BOM Effective Date Management
I am hoping that one or more of you have a better way to manage
engineering changes that require the setting and using of
effective/discontinued dates in bills of material.
We are on BPCS version 8.0 (and planning conversion to LX).
Our practice (on engineering changes for use-up of the old part)
is to:
1 - Estimate the day on which we will run out of the old part,
and enter that discontinued date into all the current bills of material
that call for the part. (After making sure that this date will give us
enough time to get the new part).
2 - Add the new part to the same BOM, effective the next day.
3 - Later, when shop orders are issued, check for new and old
parts, and if the estimate was off, change the shop order BOM to reflect
the correct part number and adjust the effective date. (Of course we
wait until approximately the estimated effective date to do this check).
The situation is complicated by the fact that we are running
behind our planned schedules - that is, some new shop orders are being
issued later than their original planned starting date, some earlier.
One reason for this is that we increasingly are checking to make sure
that all components are on hand before starting a job - if they are not,
we delay start and expedite - but this combines with the effective date
management problem to result in often checking for the wrong part (We
determine that part A is here, but really we should have been checking
for part B).
Obviously, getting all jobs to schedule would greatly simplify
the problem, and we are working on that ( with some success).
Can anyone suggest improvement solutions?
Dick Bailey
MCFA, Inc.
midrange.com
