Consolidating demand will help to prevent being over-stocked or under-stocked.
What things need to be taken into account when we consolidate:
Probability of occurrence?
An example for forecasts without consolidation:
dynBER forecast: 1 IDG in 8 months SV probability 90%
Pool forecast: 1 IDG in 10 months SV probability 20%
Disassembly: 1 IDG in 6 months AR probability 64%
Without consolidation we would just buy 3 IDGs.
But do we really need 3 IDGs?
Also can you combine all types of demand. For instance, could you combine these two forecasts?
Pool Forecast 1 IDG in 2 months SV 80% probability
Disassembly 1 IDG in 8 months AR 30% probability
Most likely not. Not only are the conditions not the same but neither is the demand owner, and lead date. So for consolidation to work there would need to be rules.
What are the possible rules for consolidation?
In the first brainstorm, we discussed the following:
POOL demand will always be needed SV. Can be combined with AR demand depending upon date needed.
DYNBER demand will always be needed SV. Can be combined with AR demand depending upon date needed.
DISASSEMBLY demand will mostly be needed in AR condition. Can be combined with SV demand, as long as the disposition reserves AR material.
TRADE demand ??
Questions for dispostion
One of the goals of the dispositon tool is to have a working labeling function/reservation system. In theory if we have material in inventory, we could reserve that material for a specific demand owner because of a forecasted need for the material.
But how would that work if forecasts have been consolidated?
Example: In inventory: 1 IDG SV condition
Reserved for this consolidated demand:
1 IDG needed in 8 months 100% probability
details of the consolidated demand:
dynBER: 1 SV IDG 50% probability in 8months
Pool: 1 SV IDG 50% probability in 10months
Now, we get a material request:
disassembly 1 IDG SV 15 days
Does reservation stay? It says we have 100% probability in 8months but really its only 50%
Input for consolidation
Image from brainstorm on forecasts
What is the input to consolidation?
We want Pool demand to mirror the planning tool that the pool planners' use (DLP)
They plan on a disposet level
This was always a flaw of the current system: the forecast would calculate on a disposet level and then convert to a PN level for consolidation
Can all forecasts work on a disposet level?
Output of forecasts:
This is necessary to figure out for consolidation of all demand forecasts
Indicates action item
Image from brainstorm on forecasts, shows the complexity of a disposet
The box shows how demand could look for one disposet could look
Box represents the disposet
The inside squares represent PNs inside the disposet
In this example:
We forecast 2 of an IDG disposet needed in 12 months The disposet contains 4 PNs 2 boxes (objects) are created, each with 4 PNs
How do we avoid being over-covered?
How might Leading PNs (LP) be factored in here? Weighting them differently.
Thoughts from this session:
Is the leading PN always the newest Mod? How is this decided?
Consolidate for the purpose of disposition-then, if there isn't enough material available to meet the consolidated demand, revert or reconsolidate for the purchase decision
Requirements of all system forecasts: disposet and probability
When we have a reactive request (real need for material by X time frame)
Z2 Auction and Single sourcing will be handled with IK (disposition)