Tripping
Use
Tripping checks whether there are systematic deviations of demand history in the forecast for the current forecast model, meaning whether the forecast values are permanently above or below the values of the demand history. If this is the case, the system initializes the base values and the mean absolute deviation (MAD) for a location product, and restricts the number of historical periods that it uses for forecast calculation. The system can thus react to systematic changes in the demand history.
The sales behavior of product A rose steeply within a short period of time due to changes in the law. You now sell on average 500 of product A per month instead of 10. In this case, tripping causes the system to only consider the demand history from the time of the change in law for the forecast calculation, instead of using the whole demand history.
The system uses tripping for the following forecast models:
· First order exponential smoothing
· Second order exponential smoothing
Features
Tripping analyzes whether or not demand, order item, and average demand size per order item lay within the forecast limits. The system calculates these limits differently for the demand, order items, and average demand size per order item.
· The system calculates the limit for demand as follows:
Upper limit = Forecast plus k times the standard deviation for the forecast
Lower limit = Forecast minus k times the standard deviation for the forecast
You can enter k on the SAP Easy Access screen under Advanced Planning and Optimization → Service Parts Planning → Planning → Forecast → Forecast Profile on the Model Evaluation tab page in the Trip Limit Parameter for Demand.
· The system defines the limit for order item according to tables that assign each interval of forecast values to a value for the upper limit and a value for the lower limit.
· The system calculates the limit for average demand size per order item as follows:
Upper limit = Forecast plus k times the standard deviation for the demand/item
Lower limit = Forecast minus k times the standard deviation for the demand/item
You can enter k in the forecast profile on the Model Evaluation in the Trip Limit Parameter for Demand/Item.
If the actual demand, actual order item, or actual demand/item lie outside of these limits, the system increases or decreases the trip counter. If a later forecast lies within the borders, the system sets the trip counter back to zero. If the counter is positive, that is, greater than zero, and the actual demand, actual order item, or actual demand/item is lower than the lower limit, the system sets the trip counter to minus one. If the counter is negative, that is, less than zero, and the actual demand, actual order item, or actual demand/item is greater than the upper limit, the system sets the counter to plus one.
If the trip counter goes above or below a positive or negative limit value, there is a trip.
· You can define the limit values for the trip counter for demand in the Tripping: Lower Limit for Demand and Tripping: Upper Limit for Demand parameters.
· You can define the limit values for the trip counter for order items in the Tripping: Lower Limit for Item and Tripping: Upper Limit for Item parameters.
· You can define the limit values for the trip counter for average demand size per order item in the Tripping: Lower Limit for Demand/Item and Tripping: Upper Limit for Demand/Item parameters.
After a trip has occurred, the system reinitializes the corresponding forecast model.
The system performs this initialization according to the weighting table that you defined in Customizing under Initialize Base Values. For more information, see the Implementation Guide (IMG) for Advanced Planning and Optimization under Supply Chain Planning → Service Parts Planning (SPP) → Forecasting.
You can see the changed values in the forecast profile on the Initialization tab page in the Base Value for Demand, Base Value for Item, Base Value for Demand/Item, MAD for Demand, MAD for Item, MAD for Demand/Item, Initialization Date for Demand, Initialization Date for Item, and Initialization Date for Demand/Item parameters.
For more information, see Model Initialization of First and Second Order Smoothing and under Model Initialization of Intermittent Model.
When the system recalculates the base values, it then sets the trip counter to zero.