Simulating the future
- From: Saulius.Maciulis@xxxxxxxxx
- Date: 13 Sep 2005 14:49:26 -0700
A company sells various computer devices with three year warranty. By
contract, a company has to repair a broken device in two days after the
client had reported that. Usually, repairing is done by displacing a
broken unit with the new one. A new unit is ordered from manufacturer
and arrives in five days. So, to fullfil the contract and to avoid any
financial penalty, a company has to keep some units in the warehouse.
Keeping too much of them is not financially effective, because after
three years they are worthless.
Given a simple table with the unit name and the number of failures
during last quarter (90 days):
UNIT NAME FAILURE#
....
KYBD W/POINT STICK 4
KYBD EASY ACCESS-INT 3
KEYBOARD,SPACESAVER,OPAL-INTL 9
KEYBOARD 7
DVD-ROM Unit internal Armada M700 6
DRV,HD,72GB,WU3,HTPLG 4
DRV,HD,40GB,U100 8
DRV,HD,36GB,1",15K,UN 14
DRV,HD,36G,WU3,1"NON-HTPLG 9
....
We need to calculate the probability of a failure of particular unit on
any accidental day. After that, we need to simulate the next quarter
and to forecast, how many times each unit will have broken and when.
And there is more, but I believe if we could somehow simulate the next
quarter then we could solve all of them.
It looks like they have missed something (for example, the number of
sold devices), but they really don't. We're not searching for answer
here, what we're searching here is any links or information which could
help to solve this exercice. Maybe you could tell the name of that
"method" or "thing" which is ussually used to solve such kind of
exercices.
We hope this is not a tough one for a statistics guy.
.
- Follow-Ups:
- Re: Simulating the future
- From: Richard Ulrich
- Re: Simulating the future
- Prev by Date: Re: SAS - exact CI for risk difference
- Next by Date: how to compare the similarity of two groups?
- Previous by thread: sum of coefficients in PLUM
- Next by thread: Re: Simulating the future
- Index(es):
Relevant Pages
|
