APO-SNP layer-to-layre optimisation

[Former Member]

Hello gurus,

Could you answer...

Is it possible to configure in APO-SNP layer-to-layre optimisation

As, for example, in the SCP (i2) or Enterprise Optimizer (River Logic)?

Thank you in advance,

Svetlana.

[frank_horlacher · Accepted solution]

Yes, it is possible with SNP.

please have a look at the documentation:

Prioritization of Customer Demands and Demands with High Volume

http://help.sap.com/saphelp_scm70/helpdata/en/21/186b398c15ef59e10000000a114084/frameset.htm

this documentation outlines the settings you need to have to perform the layered optimization.

with transaction /SAPAPO/SNP_PENGRP - Define SNP Penalty Cost Group Profiles

you can define the layers.

I add the definition of the penalty cost group at the end.

with the penalty cost groups you can group demands into layers which are handled separately in the SNP Optimizer.

This is the business process behind:

1. The
   objective of this process is to maximize the service level and the
   profitability. The first thing to take in consideration is how to manage the
   service levels, mainly for the strategic customers. The planning process needs
   to be executed in a hierarchical sequence
2. nExample
   of hierarchical process:

• Service level commitment for strategic
  customers independently of the profitability. The planning process will
  generate a plan that attempts to minimize violations of these objectives.

1. nFor
   instance, a sample hierarchy of objectives would be:
2. nPlan
   to satisfy high-priority demands as much as possible.
3. nProfit
   Optimize considering all constrains for the same priority
4. nOptimize
   the benefit for the low priority demand
5. nAt
   each level of the hierarchy, The planning process looks at the whole supply
   chain and tries to determine a plan that satisfies all the hard constraints
   (e.g. lead times, capacities, etc.) while optimizing the objective of that
   level. Then, it moves to the next level where it solves the next planning
   layer, but this time making sure that the solution obtained does not compromise
   any of the previous objectives.
6. nAt
   level 1, the planning will try to find a plan (Plan 1) that satisfies
   high-priority demands as much as possible.
7. nAt
   level 2, the planning will try to find to a plan (Plan 2) that satisfies
   low-priority demands as much as possible. This time, however, one of the
   constraints is not to compromise the first business objective. In other words,
   Plan 2 will not sacrifice high priority demands to satisfy low priority
   demands.
8. nEtc.
9. nAt
   the end of this effort, we have a constrained plan that satisfies all hard
   constraints and optimizes business objectives while respecting their relative
   importance. Because of the way it was generated, the final constrained plan
   will always be capacity-feasible i.e. without any violations of capacity
   constraints; however, it will contain demand problems (forecasts not
   satisfied).

Penalty Cost Group

Grouping criteria for demand, which allows penalty costs to be assigned.

Use

You can use penalty cost groups to prioritize sales orders for
optimization-based planning in Supply Network Planning (SNP) by assigning
customer-dependent penalty costs to them. You first create penalty cost groups
in the application-specific master data under Define Penalty Cost
Groups. You can use the Business Add-in (BAdI) /SAPAPO/SDP_RELDATA to
assign the penalty cost groups to forecasts and sales orders. Then you release
the forecasts with descriptive characteristics, such as customer, to SNP or
transfer sales orders from SAP ERP to SAP SCM via the Core Interface (CIF).

On the SNP 1 tab page of the product master data, you can assign
penalty costs for non-delivery and delays to the penalty cost group. In this
way, the penalty costs become customer-dependent, for example.

You can use the penalty cost group profile to define that the SNP Optimizer
first covers demand from penalty cost groups with a higher priority and only
then covers demand from penalty cost groups with lower priority.

The penalty cost group 0000 is the default penalty cost group. If no
penalty cost values have been defined for a penalty cost group, the system uses
the values from the default penalty cost group.

[Former Member · Accepted solution]

Hello, Frank,
Many thanks for your recommendations. It's very helpful for me.
It gives some more opportunities for planning. But for the other hand the question arises as how can to define the amount of penalty costs, if I have a lot of products?

In my business case I want to realize more flexible planning model, for example, to include the layer of planning  such as minimize backlogged demand to promise date; minimize maximum safety stock violations and minimize minimum safety stock violation, and the last profit optimization, or some combinations of that.

[frank_horlacher · Accepted solution]

I know that it is not easy to define the penalty costs.

but within each layer you can define the cost profile.

S_AP9_75000101 - Define SNP Cost Profiles

in your example you want to minimize the backlog, then you would give a higher Cost Multiplier for Late Delivery

or Penalty for Not Enough Safety Stock

and so on.

I hope this helps.

BR Frank

[Former Member · Accepted solution]

Hello Frank,
I see your point.
I was looked the tcode S_AP9_75000101 - Define SNP Cost Profiles and some description about it.
As I understand in this profile I can define the weighting factors of linear function in the process optimization. But I can use this weighting factors only in the test system. In the productive system, I should not change the standard setting. And of course I must to define penelty costs  for all layers optimisation.

In my case I want to limit linear function step by step or layer by layer.

For example, it will be in following sequence:

1 layer: Maximize demand satisfied
2 layer: Minimize backlogged demand to promise date
3 layer: Minimize minimum safety stock violations
4 layer: Minimize maximum safety stock violations for product group 1
5 layer: Minimize maximum safety stock violations for product group 2
6 layer: Maximize Total Profit

[frank_horlacher · Accepted solution]

I understand.

there are other possibilities to solve your requirement.

How do you define your layers?

If you are able to define the layers by demand grouping you can release the demands from demand planning layer by layer.

/SAPAPO/MC90 - Release Demand Planning to Supply Network Planning

choose Orders with Specific characteristics

release the demand  layer by layer to SNP.

optimize layer by layer in SNP with the setting in SNP Optimization Profile

on tab Integration

Do not delete any orders

then the result from the first layer will not be deleted.

The next layer will then be optimized with new settings in cost profiles.

and so on.

[Former Member · Accepted solution]

Hello Frank,

As I understand you suggest me to define different layers of demand, am I right?

But how can I define the different layers of safety stock or other layers?

Do you mean that if I would define some demand groups and release it sequential,

run optimization, all the result of sequential running optimization will include overall result of optimization, am I right?

When I run optimization this optimization always use objectives: Cost minimization or Profit optimization.

But, for example, when I optimize safety stock or maximize demand satisfied I can ‘t optimize this objectives!

[frank_horlacher · Accepted solution]

ok, if you have these requirements then the layered optimization with the SNP Penalty Cost Group Profiles is the only way I know to achieve the solution.

You can also contact SAP consulting to get more information and help yoiu to work out a solution for you and your company.

BR Frank

[Former Member · Accepted solution]

Hello Frank,

Thanks for you advices! It's very helpful to understand the objectives.

Best Regards,

Svetlana