Aggregated Lifecycle Planning enables you to use assignments to like profiles at the detail level when planning at aggregated level, for instance product group level. This means that the historical data for the selection at aggregate level is a closer representation of the facts and consequently allows a more accurate forecast.
This function is part of the Interchangeability suite of functions in Demand Planning. See also Product Interchangeability in Demand Planning.
Aggregated lifecycle planning makes use of aggregation. Therefore the key figures used for historical data and the forecast must sum the details at higher levels. In other words the calculation type N – No Calculation must not be used. Similarly you cannot use calculation types that average the data – types A and D.
To use aggregated lifecycle planning, you must set the Aggregated Lifecycle Planning with Like Profiles indicator in the basic settings for lifecycle planning for the relevant planning area.
In both forms of aggregated lifecycle planning (with like modeling and phase-in/out) the system creates an internal table for processing the characteristic value combinations. If this table becomes too large, the system terminates. To prevent this happening the number of characteristic value combinations as defined by the characteristics in the basic settings is checked. If the number of characteristic value combinations or the number of periods is too large, aggregated lifecycle planning is not carried out and an error message is issued. It is possible to change the system defaults with report program /SAPAPO/ RMDP_FCST_AGGLC_CUST.
The system carries out the following activities:
1. It reads any existing historical data at aggregate level.
2. It reads the selection and determines the individual characteristic value combinations at the level of the basic settings.
3. It determines for which characteristic value combinations like profiles exist
4. It determines the new aggregated historical value by taking the value from step 1 and for each detail level that has been assigned a like profile subtracting any historical value and adding the value determined using the like profile.
5. It now uses this aggregated historical data to carry out a forecast.
6. The result is initially written to the aggregate level and then disaggregated to the detail levels using the calculation type specified for the key figure in the planning area.
This last step can lead to results that you were not expecting:
· Aggregated lifecycle planning does not determine how the forecast value is disaggregated. You specify the calculation type for a key figure in planning area maintenance. See Aggregation and Disaggregation.
· If you are using a Pro Rata calculation type and there are already values in some cells and not in others, only those cells with values receive a proportion of the forecast value.
Tip: If you are executing a forecast again, for instance with a modified forecast model, first delete the old forecast values before starting the forecast. (You may want to copy the old forecast values to another key figure or version.)
· If you are disaggregating based on another key figure, calculation type P, remember that may be not be any values for the new characteristic value in this key figure. If there are not, no values will appear at detail level. This means that before you carry out lifecycle planning you must enter values for the new characteristic value in the key figure on which the disaggregation of the forecast key figure is based. This is particularly important if you base disaggregation on the APODPDANT key figure. Here you cannot use proportional factor calculation function since there are no other key figure values for the new characteristic value combination on which to base the calculation. In this case you can for example copy proportional factors manually or by using a macro. To do so you must have set the Maint. props manually (Maintain proportional factors manually) indicator for the planning book and assigned the key figure to a data view.
· If you have included a characteristic value that neither has historical data not has been assigned a like profile and you are using pro rata disaggregation, a value can still be disaggregated to the forecast key figure.
A log is written during forecasting, both in interactive planning (on the Messages tab page) and in background planning. This log tells you how many characteristic value combinations have been assigned like profiles and displays the names of the like profiles used.
This function does not work with historical data that is stored in InfoCubes.
You cannot work with like profiles that contain a lag.
To prevent memory overflow which causes a program termination, the system checks the number of characteristic combinations at detail level and the number of periods for which aggregated lifecycle planning. If one of these parameters is exceeded, aggregated lifecycle planning is stopped and a message is issued. You can change these settings with the report program /SAPAPO/RMDP_FCST_AGGL_CUST.