Forecasting in SAP

™™

FORCASTING (LO-PR)FORCASTING (LO-PR)

Material ForecastingMaterial Forecasting

Forecasting in SAPForecasting in SAP

(LO-PR)(LO-PR)

22

™™


Systems Engineering, King Fahd University of Petroleum & MineralsSystems Engineering, King Fahd University of Petroleum & Minerals Material ForecastingMaterial Forecasting

Objective of the LAB sessionsObjective of the LAB sessions1.1. After these lab After these lab

sessions you will sessions you will be able to do be able to do

‘‘forecasting in forecasting in SAP’SAP’

2.2. You will be able to You will be able to put into practice put into practice the concepts learnt the concepts learnt in the previous in the previous sessionssessions

33

™™



Windows & SAP Logon InformationWindows & SAP Logon Information

Windows LogonWindows Logon information is as follows: information is as follows: Windows UserID is ‘Windows UserID is ‘sesapsesap’ & Password is ’ & Password is

‘‘sap$2005sap$2005’’

SAP logon SAP logon information is as follows:information is as follows: SAP LogonSAP Logon shortcut is available on the Desktop … shortcut is available on the Desktop … Client 800Client 800 user ID is ‘user ID is ‘aramcoxx’aramcoxx’ where xx: where xx: 01, 02, 03 - 1501, 02, 03 - 15 Your Password is ‘Your Password is ‘enter197’enter197’

™™



Material ForecastMaterial Forecast

(LO-PR)(LO-PR)

55

™™



Material ForecastMaterial Forecast

The The results of the material forecastresults of the material forecast are are frequently usedfrequently used::• as a as a basis for SOPbasis for SOP. . • In MRP, the forecast values calculated during the forecast In MRP, the forecast values calculated during the forecast

run run result in purchase requisitionsresult in purchase requisitions • or in or in planned ordersplanned orders • and they can also be used for and they can also be used for determining safety stock determining safety stock

and reorder levelsand reorder levels..

66

™™



Forecasting in SAP™:Forecasting in SAP™: Step – by – StepStep – by – Step

Available Forecast ModelsAvailable Forecast Models Forecast ParametersForecast Parameters Forecast Model SelectionForecast Model Selection

Manual Model SelectionManual Model Selection Automatic Model SelectionAutomatic Model Selection Manual Model Selection with Additional System CheckManual Model Selection with Additional System Check Model InitializationModel Initialization

Ex-Post ForecastEx-Post Forecast Parameter OptimizationParameter Optimization Maintaining Forecast ParametersMaintaining Forecast Parameters Forecast Parameters: Independent of the Forecast ModelForecast Parameters: Independent of the Forecast Model Forecast Parameters: Dependent on the Forecast ModelForecast Parameters: Dependent on the Forecast Model Creating Historical DataCreating Historical Data Creating Forecast ValuesCreating Forecast Values Carrying out the Forecast with Reference to Another MaterialCarrying out the Forecast with Reference to Another Material

77

™™



Forecast ModelsForecast Models

-------- Already covered earlier ---------------- Already covered earlier ---------------- Summary ---------------- Summary --------Forecast ModelsForecast Models

Models Models

• ConstantConstant• TrendTrend• SeasonalSeasonal• Seasonal trendSeasonal trend

88

™™



Forecast Model SelectionForecast Model Selection

Before you run a forecast, you must specify which model Before you run a forecast, you must specify which model the system should use to calculate the forecast values.the system should use to calculate the forecast values.

There are three possibilities:There are three possibilities:

-------- Already covered earlier ---------------- Already covered earlier -------- Manual model selectionManual model selection Automatic model selectionAutomatic model selection

-------- Already covered earlier ---------------- Already covered earlier --------

Manual Model Selection with Additional System Check Manual Model Selection with Additional System Check

99

™™




-------- Already covered earlier ---------------- Already covered earlier ---------------- Summary ---------------- Summary --------

1010

™™




-------- Summary Cont. ---------------- Summary Cont. --------

Manual Model SelectionManual Model Selection If you want to select a model manually, you must first analyze the historical data to determine If you want to select a model manually, you must first analyze the historical data to determine

whether a distinct pattern or trend exists. You then define your forecast model accordingly.whether a distinct pattern or trend exists. You then define your forecast model accordingly.

1111

™™





Automatic Model Selection – Automatic Model Selection – Model Selection Procedure: Model Selection Procedure: If the If the system is to make an automatic model selection, you have system is to make an automatic model selection, you have the choice between two model selection procedures:the choice between two model selection procedures:

Procedure 1: Procedure 1: The system carries out statistical tests and checks whether a The system carries out statistical tests and checks whether a

trend or a seasonal requirements pattern applies.trend or a seasonal requirements pattern applies. In the trend test, the system subjects the historical values to a In the trend test, the system subjects the historical values to a

regression analysis and checks to see whether there is a regression analysis and checks to see whether there is a significant trend pattern.significant trend pattern.

In the seasonal test, the system clears the historical values of In the seasonal test, the system clears the historical values of any possible trends and carries out an autocorrelation test.any possible trends and carries out an autocorrelation test.

1212

™™





Procedure 2: Procedure 2: The system calculates the models to be tested using various The system calculates the models to be tested using various

combinations for alpha, beta and gamma. The smoothing combinations for alpha, beta and gamma. The smoothing factors are also varied between 0.2 to 0.8 in intervals of 0.2.factors are also varied between 0.2 to 0.8 in intervals of 0.2.

The model which is then chosen is the the model which The model which is then chosen is the the model which displays the lowest mean absolute deviation (MAD).displays the lowest mean absolute deviation (MAD).

Procedure 2 is more precise than procedure 1 but takes Procedure 2 is more precise than procedure 1 but takes considerably longer.considerably longer.

1313

™™




Manual Model Selection with Additional System CheckManual Model Selection with Additional System Check In the third case, you can have a combination of manual and In the third case, you can have a combination of manual and

automatic model selection: automatic model selection: you specify a model and set the system so that it additionally you specify a model and set the system so that it additionally

checks historical values for a seasonal pattern or trend.checks historical values for a seasonal pattern or trend.

Possible combinations that you can choose are as follows:Possible combinations that you can choose are as follows:1.1. You enter a trend model for the forecast model and then let You enter a trend model for the forecast model and then let

the system check historical data for a seasonal pattern by the system check historical data for a seasonal pattern by setting the model selection to seasonal for the test. setting the model selection to seasonal for the test.

2.2. You enter a seasonal model for the forecast model and then You enter a seasonal model for the forecast model and then let the system check historical data for a trend by setting the let the system check historical data for a trend by setting the model selection to trend for the test.model selection to trend for the test.

A reference is given to the forecast formulae used by the A reference is given to the forecast formulae used by the system for the individual models in Forecast Formulae.system for the individual models in Forecast Formulae.

1414

™™



Forecast Model InitializationForecast Model Initialization

-------- Already covered earlier ---------------- Already covered earlier ---------------- Summary ---------------- Summary --------

Model InitializationModel Initialization Model initialization is the process of determining the necessary Model initialization is the process of determining the necessary

model parameters -- such as the basic value, the trend value, model parameters -- such as the basic value, the trend value, and the seasonal indices -- for the selected forecast model.and the seasonal indices -- for the selected forecast model.

Initialization takes place each time a planning time series is Initialization takes place each time a planning time series is forecasted.forecasted.

The following table shows you which model parameters are The following table shows you which model parameters are necessary for each forecast model.necessary for each forecast model.

1515

™™



Forecast Model InitializationForecast Model Initialization


Model InitializationModel Initialization

As a general rule, the forecast model is initialized As a general rule, the forecast model is initialized automatically. In order to do this, the system requires a automatically. In order to do this, the system requires a certain number of historical values. This number certain number of historical values. This number depends on the forecast model, as shown in the depends on the forecast model, as shown in the following table.following table.

1616

™™



Ex-Post ForecastEx-Post Forecast

-------- Already covered earlier ---------------- Already covered earlier ---------------- Summary ---------------- Summary -------- If more historical values are available than are required for the system If more historical values are available than are required for the system

needs to initialize the model, an ex-post forecast is carried out needs to initialize the model, an ex-post forecast is carried out automatically as follows:automatically as follows:

The historical values are divided into two groups: the first group with The historical values are divided into two groups: the first group with the older values is used for initialization; the older values is used for initialization;

The more recent values in the second group are used to carry out an The more recent values in the second group are used to carry out an ex-post forecast (see the figure below).ex-post forecast (see the figure below).

1717

™™



Parameter Optimization Parameter Optimization

For forecast models with exponential smoothing, you For forecast models with exponential smoothing, you can instruct the system to can instruct the system to optimize the smoothing optimize the smoothing factorsfactors (see Forecast Parameters: Dependent on the (see Forecast Parameters: Dependent on the Forecast Model).Forecast Model).

If the system is instructed to optimize the smoothing If the system is instructed to optimize the smoothing factors, it calculates several parameter combinations factors, it calculates several parameter combinations and then selects the combination with the lowest mean and then selects the combination with the lowest mean absolute deviation (MAD).absolute deviation (MAD).

The finer The finer the degree of optimizationthe degree of optimization, the more exact, , the more exact, but also the more time-consuming, the parameter but also the more time-consuming, the parameter optimization. optimization.

You can vary the increment (also known as the degree You can vary the increment (also known as the degree of optimization) from 0.1 (low), 0.2 (medium) and 0.3 of optimization) from 0.1 (low), 0.2 (medium) and 0.3 (high).(high).

1818

™™



Parameter Optimization Parameter Optimization

If you have set the system correctly, If you have set the system correctly, it will carry out it will carry out parameter optimization during initializationparameter optimization during initialization for the initial for the initial forecast as well as for all other forecasts. forecast as well as for all other forecasts.

The most effective combination of the smoothing factors The most effective combination of the smoothing factors is determined by means of the ex-post forecast.is determined by means of the ex-post forecast.

Please refer to reference document for additional Please refer to reference document for additional information on the calculation of the MAD as well as the information on the calculation of the MAD as well as the significance of the smoothing factors.significance of the smoothing factors.

Optimizing the smoothing factor Optimizing the smoothing factor is only possibleis only possible for for models with the exponential smoothing procedure.models with the exponential smoothing procedure.

1919

™™



Maintaining Forecast ParametersMaintaining Forecast Parameters

You maintain forecast parameters in the material master You maintain forecast parameters in the material master record. In order to do this, proceed as is described below:record. In order to do this, proceed as is described below:

• Starting from the menu screen of the material master Starting from the menu screen of the material master record, select record, select Material Material Create general. Create general.

• Enter the material type, allocate it to an industry sector Enter the material type, allocate it to an industry sector and press ENTER . and press ENTER .

• The pop-up window Select View(s) appears on your screen.The pop-up window Select View(s) appears on your screen.• Select the Forecasting view and press ENTER . Select the Forecasting view and press ENTER . • The Organizational Levels/Profiles pop-up window now The Organizational Levels/Profiles pop-up window now

appears.appears.

2020

™™




• Because material requirements are forecast at plant level, Because material requirements are forecast at plant level, you must enter the appropriate plant you must enter the appropriate plant

• Press ENTER . Press ENTER . • The "Forecasting" data screen now appears.The "Forecasting" data screen now appears.• Fill in the data screen and save your entries via Material Fill in the data screen and save your entries via Material

Save.Save.

A more detailed description is given of the individual forecast A more detailed description is given of the individual forecast parameters in the following sections:parameters in the following sections:

• Forecast Parameters: Independent of the Forecast ModelForecast Parameters: Independent of the Forecast Model• Forecast Parameters: Dependent on the Forecast ModelForecast Parameters: Dependent on the Forecast Model

2121

™™



Forecast Parameters: Dependent on the Forecast Forecast Parameters: Dependent on the Forecast

ModelModel

You can/must maintain the following parameters depending on You can/must maintain the following parameters depending on how you carry out model selection and which model you how you carry out model selection and which model you choose. The relationship between parameters and model choose. The relationship between parameters and model selection is shown in the following two tables.selection is shown in the following two tables.

2222

™™



Forecast Parameters: Dependent on the Forecast Forecast Parameters: Dependent on the Forecast

ModelModel

2323

™™




Forecast model:Forecast model: You determine via the forecast model You determine via the forecast model which model the system uses as a basis when which model the system uses as a basis when calculating the forecast values. If you do not know the calculating the forecast values. If you do not know the forecast model, you can have it determined by the forecast model, you can have it determined by the system via automatic model selection. system via automatic model selection.

Model selection: Model selection: This indicator specifies for which model This indicator specifies for which model the system is to examine the historical values for any the system is to examine the historical values for any selected model. selected model.

• Please note that depending on the model test, a minimum Please note that depending on the model test, a minimum number of historical values must be available. This field is number of historical values must be available. This field is significant if you do not know the model and you want the significant if you do not know the model and you want the system to determine it automatically. Furthermore, you system to determine it automatically. Furthermore, you also have the possibility of pre-selecting a trend model, but also have the possibility of pre-selecting a trend model, but at the same time instruct the system to search for a at the same time instruct the system to search for a seasonal pattern and vice versa.seasonal pattern and vice versa.

2424

™™




Selection procedure: Selection procedure: This indicator specifies how the system is This indicator specifies how the system is to carry out the model selection. Here, you can choose to carry out the model selection. Here, you can choose between two procedures:between two procedures:

Parameter optimization: Parameter optimization: Via this indicator, you can specify Via this indicator, you can specify that the system is to optimize the necessary smoothing that the system is to optimize the necessary smoothing factors for the appropriate model. factors for the appropriate model.

Periods per seasonal cycle: Periods per seasonal cycle: You must enter the number of You must enter the number of periods that constitute a season here if you have selected a periods that constitute a season here if you have selected a seasonal model or if the system is to carry out a seasonal test. seasonal model or if the system is to carry out a seasonal test.

Optimization level: Optimization level: By determining the optimization level, you By determining the optimization level, you are specifying the increment with which the system is to carry are specifying the increment with which the system is to carry out parameter optimization. out parameter optimization.

2525

™™




Weighting group: Weighting group: You only have to maintain this field if You only have to maintain this field if you selected the forecast model, "weighted moving you selected the forecast model, "weighted moving average". This key specifies how many historical values average". This key specifies how many historical values are taken into account for the forecast and how these are taken into account for the forecast and how these values are weighted in the forecast calculation.values are weighted in the forecast calculation.

The following factors are used by the system, depending The following factors are used by the system, depending on the model, for exponential smoothing. on the model, for exponential smoothing.

• Alpha factorAlpha factor• Beta factorBeta factor• Gamma factorGamma factor• Delta factorDelta factor

If you set parameter optimization, the system will If you set parameter optimization, the system will overwrite the originally set smoothing factors with those overwrite the originally set smoothing factors with those which have been newly calculated by the optimization which have been newly calculated by the optimization process.process.

2626

™™



Creating Historical Data Creating Historical Data

Historical data is generally updated in the material master record Historical data is generally updated in the material master record when stock is withdrawn from the warehouse. However, it is also when stock is withdrawn from the warehouse. However, it is also possible to possible to enter historical data manuallyenter historical data manually as well. as well.

There are two types of historical data (consmp. values) in the There are two types of historical data (consmp. values) in the material master record: material master record: total consumptiontotal consumption and and unplanned unplanned consumption values.consumption values.

• Total consumption is the total of planned and unplanned consumption. Total consumption is the total of planned and unplanned consumption. For materials planned according to the consumption-based planning For materials planned according to the consumption-based planning procedure, it is always the total consumption that is updated.procedure, it is always the total consumption that is updated.

• Unplanned consumption is only updated for materials that are planned Unplanned consumption is only updated for materials that are planned according to MRP, according to MRP,

if goods are withdrawn from the warehouse without a reservationif goods are withdrawn from the warehouse without a reservation if goods are withdrawn from the warehouse on the basis of a if goods are withdrawn from the warehouse on the basis of a

reservation, but if the quantity withdrawn exceeds the reservationreservation, but if the quantity withdrawn exceeds the reservation When withdrawing a material that is planned according to the MRP When withdrawing a material that is planned according to the MRP

procedure on the basis of a reservation, only the planned procedure on the basis of a reservation, only the planned consumption is updated. consumption is updated.

2727

™™



Creating Historical Data Creating Historical Data

This is how you enter consumption values manually:This is how you enter consumption values manually:• In the Forecasting or Material Requirements Planning screen that In the Forecasting or Material Requirements Planning screen that

you have already filled in, specify whether the consumption values you have already filled in, specify whether the consumption values are managed in days, weeks, or months by filling in the Period are managed in days, weeks, or months by filling in the Period indicator field. indicator field.

• In the Forecasting screen, check the proposed values in the area In the Forecasting screen, check the proposed values in the area Number of periods required, and change or add to them as Number of periods required, and change or add to them as necessary. necessary.

• Click on ‘consumption vals’ button.Click on ‘consumption vals’ button.• The consumption values entry screen appears.The consumption values entry screen appears.• Enter your consumption values, starting from the current period Enter your consumption values, starting from the current period

and working backwards. and working backwards. • You can enter as many values as you want, both for total You can enter as many values as you want, both for total

consumption and for unplanned consumption. Switch between the consumption and for unplanned consumption. Switch between the total consumption and unplanned consumption entry screens.total consumption and unplanned consumption entry screens.

• If you want to enter more values than there is room for on the first If you want to enter more values than there is room for on the first screen, press F23 to obtain further empty lines.screen, press F23 to obtain further empty lines.

• Store the data in the system with Material Store the data in the system with Material Save. Save.

2828

™™



Creating Forecast Values Creating Forecast Values

Forecast values are generally calculated by the system as part Forecast values are generally calculated by the system as part of the forecasting process, and then updated in the material of the forecasting process, and then updated in the material master record.master record.

However, it is possible to enter forecast values manually, However, it is possible to enter forecast values manually, which then form the basis of the planning run.which then form the basis of the planning run.

This is how you enter forecast values:This is how you enter forecast values:• Once you have specified the required entries on the Forecasting Once you have specified the required entries on the Forecasting

data screen, define whether forecast values are managed in days, data screen, define whether forecast values are managed in days, weeks, or months by filling in the Period indicator field. weeks, or months by filling in the Period indicator field.

• Click ‘Forecast values’. Click ‘Forecast values’. • The forecast values entry screen appears.The forecast values entry screen appears.

• Enter your forecast values. Enter your forecast values. • You can enter as many forecast values as you want. If you want to You can enter as many forecast values as you want. If you want to

enter more values than there is room for on the screen, press F23 enter more values than there is room for on the screen, press F23 to obtain further empty linesto obtain further empty lines

2929

™™



Carrying out the Forecast with Reference to Carrying out the Forecast with Reference to Another Material Another Material

If If no historical data exists for materialsno historical data exists for materials, which the system can , which the system can access during the forecast, it is possible to use another access during the forecast, it is possible to use another material for the forecast. material for the forecast.

To do this, you must define a To do this, you must define a so-called reference materialso-called reference material for for consumption as well as a consumption as well as a reference plant, a validity date, and reference plant, a validity date, and a multipliera multiplier..

If you have defined a reference material, the system accesses If you have defined a reference material, the system accesses the historical data of the reference material during the the historical data of the reference material during the forecast up to the specified validity date. From the validity forecast up to the specified validity date. From the validity date, the system then uses the material’s own historical data. date, the system then uses the material’s own historical data.

By entering a multiplier, you can define that only a certain By entering a multiplier, you can define that only a certain percentage of the consumption quantity of the reference percentage of the consumption quantity of the reference material is used for the forecast. material is used for the forecast.

3030

™™



Maintaining Parameters Using Profiles Maintaining Parameters Using Profiles

You can maintain forecast parameters using profiles. It stores You can maintain forecast parameters using profiles. It stores standard information that you need repeatedly in almost identical standard information that you need repeatedly in almost identical combinations when maintaining different objects. combinations when maintaining different objects.

Thus, a profile simplifies data entry and data management.Thus, a profile simplifies data entry and data management. In a profile you specify;In a profile you specify;

• which fields are filled with values when data on an object is entered,which fields are filled with values when data on an object is entered,• which values these fields contain andwhich values these fields contain and• which of these values can be overwritten (default values) and which which of these values can be overwritten (default values) and which

cannot be overwritten (fixed values).cannot be overwritten (fixed values).

The following functions are available for profiles:The following functions are available for profiles:• Creating a ProfileCreating a Profile• Displaying a ProfileDisplaying a Profile• Changing a ProfileChanging a Profile• Deleting a ProfileDeleting a Profile

For more info: For more info: http://help.sap.com/saphelp_46c/helpdata/en/51/a0d53437a05b2ee10000009b38f83http://help.sap.com/saphelp_46c/helpdata/en/51/a0d53437a05b2ee10000009b38f83b/frameset.htmb/frameset.htm

™™



Carrying Out ForecastCarrying Out Forecast

(LO-PR)(LO-PR)

3232

™™



Carrying Out the ForecastCarrying Out the Forecast

OverviewOverview Things to Think About Before Carrying out the ForecastThings to Think About Before Carrying out the Forecast Forecast OptionsForecast Options Carrying Out the Individual ForecastCarrying Out the Individual Forecast Carrying Out the Total ForecastCarrying Out the Total Forecast Carrying Out the Total Forecast in Background ModeCarrying Out the Total Forecast in Background Mode Calculating the Forecast ValuesCalculating the Forecast Values Calculating the Safety Stock and the Reorder LevelCalculating the Safety Stock and the Reorder Level Ex-Post ForecastEx-Post Forecast Monitoring the Selected Forecast ModelMonitoring the Selected Forecast Model How Does the System Optimize the Smoothing FactorsHow Does the System Optimize the Smoothing Factors

3333

™™




Things to Think About Before Carrying out the Forecast Things to Think About Before Carrying out the Forecast Before you can carry out the forecast for an article, the following Before you can carry out the forecast for an article, the following

prerequisites must be fulfilled:prerequisites must be fulfilled:• You must have You must have maintained the forecast parametersmaintained the forecast parameters..• You must have determined You must have determined the periodthe period for which the system is to carry for which the system is to carry

out the forecast. You do this by defining the period indicator in the article out the forecast. You do this by defining the period indicator in the article master. master.

• For each article, the forecast can be carried out on either a daily, weekly, For each article, the forecast can be carried out on either a daily, weekly, or monthly basis, or alternatively, according to accounting period.or monthly basis, or alternatively, according to accounting period.

• You must already have determined how the system is to You must already have determined how the system is to select the select the forecast modelforecast model. This can either be done manually or automatically by the . This can either be done manually or automatically by the system. system.

• In the case of automatic model selection, you must make sure that the In the case of automatic model selection, you must make sure that the required number of historical values are available; the number required required number of historical values are available; the number required varies depending on the forecast model selected.varies depending on the forecast model selected.

• If no historical values were entered when creating the article master, you If no historical values were entered when creating the article master, you must carry out manual model initialization.must carry out manual model initialization.

3434

™™




Things to Think About Before Carrying out the Forecast Things to Think About Before Carrying out the Forecast Before carrying out the forecast, you also have the possibility Before carrying out the forecast, you also have the possibility

of revising an article's historical values which are normally of revising an article's historical values which are normally automatically updated by the system during withdrawal automatically updated by the system during withdrawal postings. On the one hand, you can achieve this by doing one postings. On the one hand, you can achieve this by doing one of the following;of the following;

• manually changing the values in the article master by means of manually changing the values in the article master by means of the correction valuesthe correction values

• correcting the weighting factor for the historical values in correcting the weighting factor for the historical values in customizing - for the weighted moving average model.customizing - for the weighted moving average model.

• On the other hand, you can reduce the number of periods to be On the other hand, you can reduce the number of periods to be taken into consideration by limiting the number of historical values taken into consideration by limiting the number of historical values . This means that not all of the historical values are used for the . This means that not all of the historical values are used for the forecast, and instead, only the number of values that you enter. forecast, and instead, only the number of values that you enter.

3535

™™




If a structural interruption appears in the time series of past If a structural interruption appears in the time series of past values stored in the article master record, and if for the next values stored in the article master record, and if for the next forecast you only want the system to take those values into forecast you only want the system to take those values into account which lie after the interruption, then you do this by account which lie after the interruption, then you do this by entering the appropriate number of past time periods in the entering the appropriate number of past time periods in the Historical values field.Historical values field.

Consumption values for the period in which the article was Consumption values for the period in which the article was created as well as in the previous period. These periods can created as well as in the previous period. These periods can only be influenced by consumption postings for the respective only be influenced by consumption postings for the respective period.period.

3636

™™



Forecast OptionsForecast Options

There are several possibilities available to you for carrying out There are several possibilities available to you for carrying out the article forecast:the article forecast:

• You can carry out the You can carry out the individual forecastindividual forecast. This means that . This means that you carry out the forecast for a certain article in a certain you carry out the forecast for a certain article in a certain site.site.

• You can carry out the You can carry out the total forecasttotal forecast. This means that you . This means that you carry out the forecast for all articles of one particular site carry out the forecast for all articles of one particular site or for all sites.or for all sites.

• To improve performance, it is advisable to carry out the To improve performance, it is advisable to carry out the total forecast in background modetotal forecast in background mode..

3737

™™



Forecast OptionsForecast Options

3838

™™



Calculating the Forecast Values Calculating the Forecast Values

The forecast calculates the forecast values by using The forecast calculates the forecast values by using various various mathematical formulaemathematical formulae (ref.) which evaluate historical data. (ref.) which evaluate historical data. The accuracy of the forecast basically depends on the extent The accuracy of the forecast basically depends on the extent and quality of the data that is available.and quality of the data that is available.

You only enter historical values when you create a new article You only enter historical values when you create a new article master record or when the system creates a new article master record or when the system creates a new article master record via batch input. The system then generates master record via batch input. The system then generates these values using withdrawal postings.these values using withdrawal postings.

3939

™™



Calculating the Forecast Values Calculating the Forecast Values

The system recalculates the basic value and the trend value The system recalculates the basic value and the trend value for each forecast by using the smoothing factors and the most for each forecast by using the smoothing factors and the most recent consumption values. recent consumption values.

The smoothing factors are the weighting factors used for The smoothing factors are the weighting factors used for historical data, with the values from the most recent period historical data, with the values from the most recent period playing a more important role in the forecast than the values playing a more important role in the forecast than the values from previous periods. from previous periods.

The greater the smoothing factors you select in the article The greater the smoothing factors you select in the article master, the faster the forecast values are adapted to the master, the faster the forecast values are adapted to the actual past consumption values. actual past consumption values.

4040

™™



Calculating the Safety Stock and the Reorder Calculating the Safety Stock and the Reorder Level Level

If a warehouse that stores article has to be ready for delivery at If a warehouse that stores article has to be ready for delivery at all times, then there has to be a high level of safety stock in this all times, then there has to be a high level of safety stock in this warehouse, warehouse, as forecasting error cannot always be ruled outas forecasting error cannot always be ruled out. .

This would also entail a high level of warehousing costs. The This would also entail a high level of warehousing costs. The level of safety stock depends on the level of safety stock depends on the service levelservice level you select, on you select, on the replenishment lead time and on the accuracy of the forecast.the replenishment lead time and on the accuracy of the forecast.

In order to keep the safety stock and therefore the storage costs In order to keep the safety stock and therefore the storage costs as low as possible, the RP controller usually specifies a service as low as possible, the RP controller usually specifies a service level for each article. level for each article.

From a mathematical point of view, the service level represents From a mathematical point of view, the service level represents an organization's ability to avoid a shortfall occurring during an organization's ability to avoid a shortfall occurring during replenishment lead time.replenishment lead time.

4141

™™



Calculating the Safety Stock and the Reorder Calculating the Safety Stock and the Reorder Level Level

The replenishment lead time refers to the in-house production The replenishment lead time refers to the in-house production time for in-house production and the planned delivery time for time for in-house production and the planned delivery time for external procurement. external procurement.

Finally, the safety stock level also depends on the accuracy of Finally, the safety stock level also depends on the accuracy of the forecast . If the forecast deviates greatly from the actual the forecast . If the forecast deviates greatly from the actual consumption values, then the safety stock level will also be consumption values, then the safety stock level will also be unusually high.unusually high.

4242

™™



Ex-Post Forecast Ex-Post Forecast

The system carries out an ex-post forecast for the initial The system carries out an ex-post forecast for the initial forecast if more past values are available than are forecast if more past values are available than are required/are to be used for initialization. required/are to be used for initialization.

It also carries out the ex-post forecast during parameter It also carries out the ex-post forecast during parameter optimization, model initialization and for evaluating the optimization, model initialization and for evaluating the accuracy of the forecast.accuracy of the forecast.

4343

™™



Monitoring the Selected Forecast Monitoring the Selected Forecast Model Model

In the course of time, the characteristics of a consumption In the course of time, the characteristics of a consumption series can changeseries can change. In order to be able to react to this in time, . In order to be able to react to this in time, the system calculates a so-called the system calculates a so-called tracking signaltracking signal during every during every forecastforecast. The tracking signal links the error total (FS) and the . The tracking signal links the error total (FS) and the mean absolute devialtion (MAD) in the following manner:mean absolute devialtion (MAD) in the following manner:

• Tracking signal = │ FS/MAD │ Tracking signal = │ FS/MAD │

If the tracking signal is greater than the tracking limit, If the tracking signal is greater than the tracking limit, the the system issues an exception messagesystem issues an exception message asking you to check the asking you to check the forecast model.forecast model.

The forecast error is the difference between actual The forecast error is the difference between actual consumption values and the forecast values from the same consumption values and the forecast values from the same period whereas the error total is the sum of all the forecast period whereas the error total is the sum of all the forecast errors in a consumption series. errors in a consumption series.

4444

™™



Monitoring the Selected Forecast Monitoring the Selected Forecast Model Model

If a model is still valid, that is, if the consumption series has If a model is still valid, that is, if the consumption series has not changed, then you can assume that the error total is not changed, then you can assume that the error total is distributed normally and has an average of zero.distributed normally and has an average of zero.

If the consumption pattern has changed, however, the error If the consumption pattern has changed, however, the error total will no longer be equal to zero. total will no longer be equal to zero.

You must standardize the calculated forecast error in order to You must standardize the calculated forecast error in order to set boundaries. Therefore, in addition to the error total, the set boundaries. Therefore, in addition to the error total, the system also calculates the mean absolute deviation (MAD) as system also calculates the mean absolute deviation (MAD) as a second value. The system then adds the errors (irrespective a second value. The system then adds the errors (irrespective of the plus or minus sign) and divides them by the number of of the plus or minus sign) and divides them by the number of consumption values. The first-order exponential smoothing consumption values. The first-order exponential smoothing procedure is used to calculate the MAD. The smoothing factor procedure is used to calculate the MAD. The smoothing factor used for this is the delta factor.used for this is the delta factor.

4545

™™



How Does the System Optimize the Smoothing How Does the System Optimize the Smoothing Factors Factors

The process of the ex-post forecast described above is used The process of the ex-post forecast described above is used for parameter optimization in order to optimize the smoothing for parameter optimization in order to optimize the smoothing factors.factors.

In the article master, you can store whether parameter In the article master, you can store whether parameter optimization is to take place during the forecast and if it is to optimization is to take place during the forecast and if it is to take place, the degree of optimization you want to select.take place, the degree of optimization you want to select.

When a forecast is carried out with parameter optimization, When a forecast is carried out with parameter optimization, the system will start from an initial value and then gradually the system will start from an initial value and then gradually increase the increment of the smoothing factors in every increase the increment of the smoothing factors in every simulation run. (This constitutes the first stage of the ex-post simulation run. (This constitutes the first stage of the ex-post forecast.) forecast.)

4646

™™



How Does the System Optimize the Smoothing How Does the System Optimize the Smoothing Factors Factors

The increment in this case is specified in the article master The increment in this case is specified in the article master record by the optimization level. record by the optimization level.

The system then further analyzes the situation for the The system then further analyzes the situation for the parameter combination with the smallest MAD (second stage). parameter combination with the smallest MAD (second stage).

The optimum parameter combination is the combination which The optimum parameter combination is the combination which has the lowest possible MAD. has the lowest possible MAD.

4747

™™



Forecast ResultForecast Result

After a forecast run, the results must be checked over to make After a forecast run, the results must be checked over to make sure that the forecast was carried out correctly. There are two sure that the forecast was carried out correctly. There are two ways of doing this:ways of doing this:

Checking the Forecast Result Using a ListChecking the Forecast Result Using a List• You can print a list of the materials which were included in the You can print a list of the materials which were included in the

forecast via a print program and then you can use this list to check forecast via a print program and then you can use this list to check the results. the results.

Checking the Forecast Result in Online ModeChecking the Forecast Result in Online Mode• Alternatively, you can check for incorrect forecast results online.Alternatively, you can check for incorrect forecast results online.

You will also receive information on Changing the Forecast You will also receive information on Changing the Forecast Result.Result.

4848

™™



Changing the Forecast Results Changing the Forecast Results

You correct the forecast results in the same way as you You correct the forecast results in the same way as you correct the historical values on which the forecast is based.correct the historical values on which the forecast is based.

• One method is to enter a new value in the Corrected One method is to enter a new value in the Corrected forecast value column of the Forecast Results dialog box.forecast value column of the Forecast Results dialog box.

• Alternatively, you can use the interactive graphicsAlternatively, you can use the interactive graphics

4949

™™



Significance of the Material Forecast for Materials Significance of the Material Forecast for Materials Planning Planning

Reducing Forecast Requirements Reducing Forecast Requirements • When calculating the current stock/requirements situation When calculating the current stock/requirements situation

(displayed in the stock/requirements list), the system reduces the (displayed in the stock/requirements list), the system reduces the current period's forecast requirements by the consumption figures current period's forecast requirements by the consumption figures from the same period which have already been posted.from the same period which have already been posted.

Effects on Consumption-Based Planning Procedures Effects on Consumption-Based Planning Procedures • A forecast run has different effects on the different MRP A forecast run has different effects on the different MRP

procedures which support the forecast:procedures which support the forecast: In the automatic reorder point planning procedure, the reorder In the automatic reorder point planning procedure, the reorder

level and the safety stock are automatically determined level and the safety stock are automatically determined during the forecast run.during the forecast run.

In forecast-based planning, the results of the forecast are In forecast-based planning, the results of the forecast are copied into material requirements planning as forecast copied into material requirements planning as forecast requirements and the safety stock level is calculated.requirements and the safety stock level is calculated.

™™



END OF LAB SESSIONSEND OF LAB SESSIONS

THANK YOUTHANK YOU

Forecasting in SAP

Documents

Transcript of Forecasting in SAP