Planning Operators

Once the inputs have been validated, the inventory optimization algorithms should be activated. Depending on the data volume and the corresponding run times, this either can happen as a batch-based process that runs overnight or can be triggered by the planner.

As you’ve learned so far, with the configuration and the details about optimization algorithms, running inventory optimization requires a sequence of steps—or planning operators—to be executed. The following figure shows the standard way inventory planning operators should be structured to get all elements calculated accurately.

The following steps are typically part of a complete inventory optimization run:

1. Calculate the forecast error CV, which we mentioned in the first step as a validation element. Typically you would include this step in the end-to-end run.
2. Run the Multi-Stage Network operator to visualize multi-stage networks that are considered in optimization (optionally).
3. Run the Supply Lead Time Operator to calculate inputs for inventory planning related to production, transportation, and/or supplier lead time (optionally).
4. Run the global (multistage) inventory optimization, which optimizes the stock globally and simultaneously over all products and locations, considering all drivers for safety stock, as described in previous chapters. The global run also makes sure that all the propagated demands are calculated, which might be required by subsequent operators.
5. Calculate target inventory components such as cycle stock, pipeline stock, merchandising stock, and the average inventory position.
6. Run the Service Level Prediction operator to determine how adjustments will impact service levels (optionally).

Optionally, you can store the previous cycle values of inventory optimization in a set of key figures with previous cycle in the name—for example, Previous Cycle Target Inventory. This allows the planner to compare the newly calculated results to the previous ones. This operator, of course, should only be executed once, at the start of the cycle.

Beside the planning operators, that are specifically designed for the usage in the environment of inventory optimization, it is also possible to utilize planning operators provided in the area of MRO (Maintenance, Repair, and Operations) as they are available for inventory users of IBP as well.

As mentioned in the beginning of this section, upon completion of the first step of validation of inputs, this job can be triggered automatically. Alternatively, planners can trigger it manually from the SAP IBP, add-in for Microsoft Excel, by pressing Run in the Advanced section, and selecting Inventory Optimization.

In the chapter three , you learned about the basic building blocks of the planning model, such as master data types, attributes, key figures, time profile, planning level, and so on. The planning model is the foundation of SAP Integrated Business Planning (IBP) for Supply Chain. All these entities together can be used to model the structure of an organization’s supply network and the material flows in the network.

In this section, we’ll cover how the basic building blocks of a planning model can be used to represent the supply network, which is built from locations, distribution centers, manufacturing facilities, customers and suppliers, products, and bills of materials.

A location is an actual, physical location in a supply network where products or resources are managed. There are two main categories of location in any supply network: distribution centers or warehouses, where a product is stored or packaged; and production plants, where a product is produced or assembled. In any organization, the number of locations or facilities depends on the level of complexity of the network and the organizations’ objectives.

For example, to have a lower response time to fulfill a customer demand, an organization would have many warehouses or distribution centers to make the product easily available for the consumer. It’s very critical for any organization to identify the location in its network that will play an inventory optimization role. For example, a cloth manufacturer that outsources all its production to contractors may want to plan inventory at the contractor locations to reduce the overall working capital. An organization may or may not want to include the cross-docking location while running inventory optimization.

In SAP IBP, a physical location is represented by the location master data type. A location is a physical place in the supply network. A location master data type has the following attributes:

• LOCID (location ID)

  The location ID is a unique identifier for a location in a supply network. This is a required field for the master data type.

• LOCREGION (location region)

  This attribute represents the region the location belongs to. For example, a location may belong to the North America or Europe region. This attribute is used mainly for reporting purposes or reviewing the data at an aggregate location region level in SAP IBP.

• LOCTYP (location type)

  This attribute represents the type of location; for example, a location may be a distribution center or a plant. This attribute is also used to differentiate a vendor or a supplier location from a distribution center or warehouse location. This attribute is mainly used for reporting purposes or reviewing the data at an aggregate level in SAP IBP.

• HOLDINGCOSTPCT (holding cost percentage)

  This attribute represents the holding cost percentage by location. This is the cost of carrying inventory expressed as a percentage by location.

• GEOLATITUDE (geographical latitude)

  This attribute represents the latitude of the address of the location. This information is used for geoanalytics in the Analytics SAP Fiori app.

• GEOLONGITUDE (geographical longitude)

  This attribute represents the longitude of the address of the location. This information is used for geoanalytics in the Analytics SAP Fiori app.

• LOCBUPAID (location business partner ID)

  This attribute represents the ID of a business partner. For example, if you’ve modeled a subcontractor in your network, this field captures the ID of the location as in your business partner’s IT system.

• LOCVALID (location validity)

  This attribute represents the validity of the location. This information is used for reporting purposes only.

A supplier or a vendor is an entity that supplies a material or service to an organization. Traditionally, a supplier is defined as an entity that supplies materials and subcontractors, and a contractor is an entity that provides both materials and services. Depending on the business, there can be many suppliers. For example, a very large manufacturing firm manufacturing an expensive product, such as Boeing, will have a large number of suppliers. Resellers such as Amazon also have a lot of suppliers.

Some suppliers are also classified as business partners. Organizations have tailored business relationships based on trust with their partners. Partners share risks with their customers, and their relationships are often fluid to accommodate any changes. Usually, key suppliers are modeled in the supply network for planning.

In SAP IBP, suppliers are modeled as locations. As described in the previous section, they are differentiated from the plants or distribution centers through the attribute location type and/or business partner ID. If the supply network ends at a supplier,

A product or material is a tangible or intangible good that is part of the business activities of a company. It’s produced, transported, stored, and sold in the supply network. Products can be of different types: finished products, raw materials, intermediates, components, packaging, and so on. Often the focus of inventory target setting will be on critical products only. Planners need to decide whether they want to plan all the products using a sophisticated inventory optimization tool or only the important ones.

In SAP IBP, a product is represented by the product master data type (PRODUCT). It represents all the materials relevant for inventory planning. The product master provides you with all the information about the products that your company purchases, manufactures, and sells. It’s the central source from which product-specific information can be retrieved. Product master data is also used in demand, supply, and S&OP planning processes. Some of its attributes are as follows:

• PRDID (product ID)

  The product ID is a unique identifier for each product.

• ABCID (ABC code)

  This attribute represents the ABC classification ID. The product’s ABC classification can be used to look up the service level.

• XYZID (XYZ code)

  This attribute represents the XYZ classification ID. The product’s XYZ classification can be used to create a differentiated process.

• PRDDESCR (product description)

  The product description gives basic, high-level information about the product.

• PRDFAMILY (product family)

  The product family is a product attribute for reporting at an aggregate level.

• BRAND (brand)

  The product brand is a product attribute for reporting at an aggregate level.

• MATTYPEID (material type ID)

  This attribute represents the material type—for example, finished good, semi-finished, or raw material. This is a product attribute for reporting at an aggregate level.

• CATEGORY

  The product category is a product attribute for reporting at an aggregate level.

A customer is a business partner with whom a business relationship exists, involving the issue of goods or services. As we move toward customer-centric supply chains, for some industries it’s of the utmost importance to model all individual customers as part of the supply network because they’re like business partners. Mostly in a B2B environment, all customers are modeled to enable differential planning. For an organization that sells its product to thousands of customers, it may not make sense to include all individual customers but instead include groups of customers.

In SAP IBP, a customer is represented by a customer master data type. A customer is modeled as an individual customer in a supply network. The customer ID in SAP IBP can correspond to, for example, the customer number, sold-to party, or ship-to party in the on-premise system. Other attributes, such as description, region code, and so on, are optional. A customer master data type has the following attributes:

• CUSTID (customer ID)

  The customer ID is a unique identifier for a customer in a supply network. This is a required field for the master data type.

• CUSTCHANNEL (channel)

  This attribute represents the channel through which a product is sold to a customer. This attribute is for reporting only.

• CUSTCOUNTRY (customer country)

  This attribute represents the country to which the customer belongs. For example, a customer may belong to a country in North America (Canada) or a country in Europe (France). This attribute is used mainly for reporting purposes or for reviewing the data at an aggregate customer region level in SAP IBP.

• CUSTGROUP (customer group)

  This attribute is used to sort customers into logical segments or groups. This attribute is required for inventory optimization to set target service levels at the group level. The groups can include industry, wholesale, retail, and so on.

• CUSTREGION (customer region)

  This attribute represents the region to which the customer belongs. For example, a customer may belong to the North America or Europe region. This attribute is mainly used for reporting purposes or reviewing the data at an aggregate region level in SAP IBP.

• CUSTSALESREP (customer sales rep)

  This attribute represents the sales rep who has a business relationship with the customer. This attribute is mainly used for reporting purposes or reviewing the data at an aggregate sales rep level.

• CUSTSALESVP (sales VP)

  This attribute represents the sales reps’ manager. This attribute is mainly used for reporting purposes or reviewing the data at an aggregate sales VP level.

• CUSTVALID (customer validity)

  This attribute represents the validity of the customer’s location. This information is used for reporting purposes only.

• GEOLATITUDE (geographical latitude)

  This attribute represents the latitude of the address of the customer. This information is used for geoanalytics in the Analytics app.

• GEOLONGITUDE (geographical longitude)

  This attribute represents the longitude of the address of the customer. This information is used for geoanalytics in the Analytics app.

• CUSTBUPAID (customer business partner ID)

  This attribute represents the ID of the business partner. For example, if you’ve modeled a customer in your network, this field captures the ID of a customer as in your business partner’s IT system.

• KEYACCOUNT (key account)

  This attribute represents whether the customer is a key account. This information is used for reporting purposes only.

Note that SAP IBP for inventory uses customer group attributes for flexibility rather than customer ID attributes. This is because SAP IBP for response and supply and/or SAP IBP for sales and operations often need to model individual customer ship-to locations as customer IDs, whereas SAP IBP for demand and other demand planning systems create forecasts at an aggregate customer group level. This flexibility supports the following three use cases:

• Demand planning forecasts and service levels are set a product-location level. In this case, all customer IDs are assigned to a "dummy" customer group.

• Demand planning forecasts and service levels are set for each individual product-location-customer ID level. In this case, the customer group attribute for each customer record can be set to be the same as the customer ID.

Source customer group master data (SOURCECUSTGROUP) represents the inventory flow to the customer group from the customer-facing stocking nodes, also referred to as the demand stream. This data is specific to a product at a customer-facing location to meet the demand for a customer group. You also define the customer-service-level-target input for each demand stream, and a record must exist for each demand stream for which there is demand key figure data (forecast or sales).

The source customer group non-root attribute is TARGETSERVICELEVEL, which represents the customer service level (between 0.5 and 1). The customer service level input is used to calculate inventory targets. The root attributes are CUSTGROUP, PRDID, and LOCID, which we defined previously.

In a supply network, a node is represented by a combination of a customer product or a location product. Because the inventory optimization model works at the customer group level, neither a customer product node nor any of its attributes are required for inventory optimization. Each location-product combination can have a different planning strategy. For example, a product at the plant may not be stored due to storage space limitation. As soon as the product is manufactured, it’s moved to a nearby warehouse in which all the products are stored. Two locations may also have a different product replenishment strategy for the same product; for example, a warehouse on the East Coast may order more frequently than a warehouse on the West Cost.

A supply network node that’s a location-product combination is represented by the location-product master data (LOCATIONPRODUCT) in SAP IBP. The attributes of the location-product combination also represent how a product’s inventory is held at a location and how the information about that product flows through that location. LOCID and PRDID were previously defined, but the other attributes are as follows:

• PLUNITID (planning unit or sub-network ID)

  The planning unit ID or sub-network ID to which this location-product is assigned, also referred to as a sub-network ID, is used to split the supply network and execute inventory operators on that limited scope. Because you’ll be executing multi-echelon inventory optimization on the network, make sure that you select the scope of the individual supply chains that are independent of each other.

• STOCKINGNODETYPE (stocking node type)

  This attribute represents the stocking node type indicator (S or N). S refers to a stocking node (inventory is held), and N refers to a non-stocking node (inventory flows through). S is the default value.

• SERVICELEVELTYPE (service level type)

  This attribute represents the service-level-type indicator (A or F). A refers to the available in full, type 1, or non-stock out-probability service level metric. F refers to fill rate or type 2. A is the default value.

• PBR (periods between review)

  The periods between review, also known as the order cycle, is the time interval (in weeks) between two consecutive order or manufacturing run decisions. It’s typically one day (1/5 or 1/7 weeks) for distribution DCs, the inverse of run frequency at plants, and the inverse of ordering frequency at procurement. It shouldn’t be based on the minimum order quantity (MOQ).

• SAFETYSTOCKPOLICY (safety stock policy)

  This attribute represents the safety stock allocation policy (F, I, or D). This attribute is relevant for how safety stock targets are calculated at hybrid stocking nodes (with both external customer demand and internal dependent demand). The options are as follows:

  • – F: Refers to first come, first served (no allocation) and implies that the total demand at the hybrid stocking node is set to the customer service level target. It’s set as the default.

  • – I: Refers to independent demand and indicates that safety stock is calculated for only the external customer demand at the customer service level target. You might use this when, for example, a plant has limited storage capacity.

  • – D: Refers to divide and indicates that safety stock is calculated separately (no risk pooling) for each demand at its own service level target.

• MAXINTERNALSERVICELEVEL (maximum internal service level)

  This attribute represents the maximum value as a constraint for the internal service level (between 0.5 and 1). A key step in the end-to-end inventory planning process is multiechelon inventory optimization (MEIO), the outcome of which is coordinated inventory targets for each node of the supply network for each period of the planning horizon. To achieve this, the MEIO solver must calculate the optimal internal service levels at each of the internal stocking points while using the customer service level targets as constraints. This attribute acts as an upper-bound constraint on the optimal internal service level.

• MININTERNALSERVICELEVEL (minimum internal service level)

  This attribute represents the minimum value as a constraint for the internal service level (between 0.5 and 1). Like MAXINTERNALSERVICELEVEL, this attribute acts as a lower-bound constraint on the optimal internal service level.

In a supply network, there are many locations that supply to another location or multiple locations. For example, a manufacturing plant supplies a product to different distribution centers. So for any given location-product combination, a product can be sourced from a plant or can be supplied to a different distribution center.

In SAP IBP, the supplying location is represented by the ship-from location master data (LOCATIONFR). This is a reference master data type that references the location master data type. It isn’t an independent master data type and doesn’t require separate data loading. It represents a sourcing location that ships to other locations. Its root attribute is LOCFR, which represents the ship-from location ID. Its values are referenced from the LOCID attribute.

Just like the ship-from location, the receiving location in SAP IBP is represented by ship-to location master data (LOCATIONTO). This is a reference master data type that references the location master data type. Like ship-from location master data, it isn’t an independent master data type and doesn’t require separate data loading. It represents a receiving location that receives the product from other locations. Its root attribute is LOCTO, which represents the ship-to location ID. Its values are referenced from the LOCID attribute.

Location sourcing rules determine how the demand is fulfilled at a location other than the customer. A location sourcing rule represents an arc that connects two locations in a supply network. Location sourcing rules are used to represent how the stock moves between different locations—for example, a stock transfer from a manufacturing location to a distribution center.

Location sourcing rules also have a location sourcing ratio, which determines the split by which the demand at the location is fulfilled from two different sourcing locations. The location sourcing ratio can be constant or time-phased. To depict a constant location sourcing ratio, the location transport ratio in the master data type can be maintained (TRATIO). If the location sourcing ratio should vary over time, the time series flag will be set in the master data type (RATIOTS = 'X'), resulting in the use of the transport ratio key figure by the inventory optimization operator.

Lead times are also available in location sourcing rules to account for the transportation time between the sending and receiving locations. Because lead time provides the only opportunity to do so, you must capture the time offset between the availability of stock at a source location and receipt of stock at the destination location.

Various factors should be considered while modeling transportation lead times. Some examples of times that may add up to overall transportation lead times are as follows:

• The time required to pick, pack, and load the product at the source location
• Physical transportation time between source and destination locations
• Other time considerations for receiving the product at the destination location, such as the time at customs

In SAP IBP, transportation sourcing information for each product, ship-from location (source), and location (target) is represented by the source location master data. SOURCELOCATION represents the PRDID, LOCFR, and LOCID root attributes, which we defined previously. The non-root attributes are as follows:

• RATIOTS (ratio time series indicator)

  This attribute represents a time-series/static indicator for sourcing quotas. This indicates if time-series values (if set to X or x) or static quota values should be used for transportation sourcing quotas.

• TRATIO (location sourcing ratio)

  This attribute represents the static transportation sourcing quota value. If RATIOTS is set to static, this value is used as the sourcing quota for all periods for the transportation lane.

• TLEADTIME (transportation lead time)

  This attribute represents the transportation lead time in weeks (can be fractional). It includes any ordering time, picking time, and so on.

• TLEADTIMEVARIABILITY (transportation lead time error CV)

  This attribute represents the transportation lead time coefficient of variation (CV) in weeks. This is the ratio of the standard deviation of lead time to the average lead time in weeks.

• TMINLOTSIZE (minimum transportation lot size)

  This attribute represents the minimum lot size for the transportation lane in the base unit of measure of the product.

• TINCLOTSIZE (incremental lot size)

  This attribute represents the incremental lot size for the transportation lane in the base unit of measure of the product.

• TSHIPMENTFREQUENCY (transportation shipment frequency)

  This attribute represents the number of shipments received per week. This is the inverse of the shipment frequency per week. This attribute is optional and should be used only in specific advanced scenarios, such as when shipments are received more frequently than the ordering frequency—for example, if orders are placed once a week but the product is shipped daily in smaller batches. This doesn’t impact safety stock requirements but does have an impact on the cycle stock (inventory to cover forecast or known net requirements). This defaults to the inverse of PBR when no value is provided (recommended).

• TDELIVERYTYPE (transportation delivery type indicator)

  This attribute represents the indicator for the impact of late deliveries. This attribute is optional and should be used only in specific advanced scenarios related to lead time variability. It specifies how two consecutive shipments are assumed to behave when the first shipment is late:

  • – S or s indicates sequential; that is, the second shipment always arrives after the first shipment (more conservative safety stock). This is the default.

  • – C or c indicates cross-over; that is, the two shipments are independent, and the second shipment can arrive before the first shipment (less conservative safety stock).

In any supply network, at a manufacturing location where a product is produced, one of the key requirements is to have a bill of materials for the product. A bill of materials is a complete, formally structured list of the components that make up a product or assembly. The list comprises the material number of each component, together with the quantity and unit of measure. A BOM can be single level or multilevel. For example, a finished material would contain semi-finished materials as components, which in turn would contain raw materials as components in the next level. There can also be different alternative BOMs for a product depending upon the lot sizes, validity dates, and different production methods.

BOMs are also used in production planning for semi-finished goods and in purchase planning for raw materials by propagating the demand through a BOM explosion.

In SAP IBP, a single bill of materials is represented by a production source header and production sourcing item master data. The production source header (SOURCEPRODUCTION) is the header information for each product, location, and source header. This can be from a manufacturing BOM or sourced from an external vendor. The source header (SOURCEID) must be unique for each product and location if it represents a BOM (has component products). PRDID and LOCID were defined previously. Scenarios in which a product can be made at a plant and simultaneously sourced from another plant are supported. The additional source production attributes are as follows:

• SOURCEID (source ID)

  This attribute represents the production source header ID. This is a unique indicator for each source production record.

• OUTPUTCOEFFICIENTTS (output product coefficient time series indicator)

  This attribute represents the output coefficient time series indicator. This is an optional attribute that indicates whether the output coefficient is a time series (X or x) or static (any other value). It defaults to static if no value is provided.

• OUTPUTCOEFFICIENT (output product coefficient)

  This attribute represents the output coefficient value. This is an optional attribute (it defaults to 1 if no value is provided) used as the output coefficient in every period if the corresponding indicator is set to static. The component coefficient in the BOM is divided by this value when determining the BOM multiplier for the components.

• PRATIO (production sourcing ratio)

  This attribute represents the static production sourcing quota value. If PRATIOTS is set to static, this value is used as the sourcing quota for all periods for the production source headers.

• PRATIOTS (production ratio time series indicator)

  This attribute represents a time series/static indicator for sourcing quotas. This indicates if time series values (if set to X or x) or static quota values should be used for production-sourcing quotas.

• PLEADTIME (production lead time)

  This attribute represents the production lead time in weeks (can be fractional). It includes any ordering time, picking time, goods issue time, and so on.

• PLEADTIMEVARIABILITY (production lead time CV)

  This attribute represents the production lead time CV in weeks. This is the ratio of the standard deviation of lead time to the average lead time in weeks.

• PMINLOTSIZE (minimum production lot size)

  This attribute represents the minimum lot size for production in the base unit of measure of the product.

• PMAXLOTSIZE (maximum production lot size)

  This attribute represents the maximum lot size for production in the base unit of measure of the product.

• PINCLOTSIZE (production incremental lot size)

  This attribute represents the incremental lot size for the production source in the base unit of measure of the product.

• PSHIPMENTFREQUENCY (production shipment frequency)

  This attribute represents the number of shipments received per week. This is the inverse of the order frequency per week. This attribute is optional and should be used only in specific advanced scenarios, such as when orders are produced more frequently than the ordering frequency—for example, if orders are placed once a week but the product is produced daily in smaller batches. This doesn’t impact safety stock requirements but does have an impact on the cycle stock (inventory to cover forecast or known net requirements). This defaults to the inverse of PBR when no value is provided (recommended).

• PDELIVERYTYPE (production delivery type)

  This attribute represents the indicator for the impact of late orders. This attribute is optional and should be used only in specific advanced scenarios related to lead time variability. It specifies how two consecutive production orders are assumed to behave when the first order is late:

  • – S or s indicates sequential; that is, the second order always finishes after the first order (more conservative safety stock). This is the default.

  • – C or c indicates cross-over; that is, the two orders are independent, and the second order can be completed before the first order (less conservative safety stock).

• SOURCETYPE (source type)

  This attribute represents the production source type indicator (P, U, or C). This dictates if the source production record represents a primary output of a BOM (P), if the product is sourced from an external vendor (U), or if it is a secondary co-product output of a BOM (C). Co-products are ignored in inventory target planning, and an external vendor represents a supply end point in the network.

Component master data (COMPONENT) is a reference master data type that references the product master data type. It represents the component products in a BOM that are used to manufacture or build or assemble the output product. Its root attribute is PRDFR, which represents the component product ID. Its values are referenced from the PRDID attribute.

Production source item master data (PRODUCTIONSOURCEITEM) represents the BOM information for each source production record of type P. PRDID and SOURCEID were defined previously. COMPONENTCOEFFICIENTTS and COMPONENTCOEFFICIENT are like OUTPUTCOEFFICIENTTS and OUTPUTCOEFFICIENT, respectively.

To configure a multi level BOM, you must use two types of master data:

• SOURCEPRODUCTION

• PRODUCTIONSOURCEITM

Production is modeled using either static or time-dependent components.

The component is declared in PRODUCTID with the quantity COMPONENTCOEFFICIENT.

The resulting product is the SOURCEITMID.

COMPONENTCOEFFICIENTTS can be set to "Y" if the coefficient is time dependent.