Effective service parts management must strike a balance between high service levels and increasing logistics costs. The key challenge is having the right part at the right time at the right location to fulfill the service requests without having huge inventory buffers or managing excess stock. Service Parts Planning is designed to help companies provide effective service parts management. Figure 1 is a chronological overview of how the planning process works in Service Parts Planning.

Figure 1

Service Parts Planning process overview for planning

The target of the planning process is to determine the required inventory levels at the customer facing locations to:

• Meet the target service level
  
  
• Plan procurement
  
  
• Plan the replenishment needs for the customer facing locations

Service Parts Planning offers a variety of planning features that support this target throughout the service parts supply chain from the appearance of the demand to the delivery of the service product. The process flow in Figure 2 shows Service Parts Planning’s planning sequence for service parts, which I’ll cover in the next five sections.

Figure 2

Planning process sequence

You can implement Service Parts Planning as an element of service parts management, or you can use a reduced system landscape. For system requirements for both options, see the sidebar “System Landscape for Service Parts Planning.”

The description of the Service Parts Planning core functionality in this article is based on SAP SCM 5.0. Additional functionality was added to Service Parts Planning in SAP SCM 2007 and SAP SCM 7.0. Table 1 provides a list of these new features.

Supply Chain Design

During supply chain design, the distribution network is modeled by bills of distribution (BODs). The BOD reflects the standard distribution path of the service part and serves as the basis of most planning activities.

A BOD always consists of one or more entry locations and can contain parent and child locations. The entry location defines the location to which the supplier delivers the parts. The location from which the customer is served is called a customer facing location. Parent locations ship to other locations that are then called children. A virtual child is introduced for locations that ship directly to the customer and also to other locations. All data related to customers, such as forecast or safety stock, is then assigned to the virtual child. As a result, stock transfers are created from the physical to the logical location to cover the demand. Figure 3 is an example of a supply chain for a service part. This supply structure transferred in a BOD results in the distribution plan shown in Figure 4.

Figure 3

Supply chain example

Figure 4

The supply structure in Figure 3 transferred in a BOD

The assignment of a product to a BOD can change over time. Data relevant to planning is then rerouted to the new BOD. The planning process also considers any future BODs. Because the BOD predefines the distribution hierarchy, the planning process can run without a time-consuming supply source determination.

The concept of virtual consolidation locations is used to group physical locations with relatively small demand into one logical location within the BOD. Although this aggregates the data for demand and replenishment planning, the physical locations are deployed according to their real demand.

Service Parts Planning also can deal with subcontractors, called contract packagers (CPs). A CP packs and repacks the goods (e.g., from containers to pallets) for the warehouses within the BOD. The CP needs to be assigned to a specific location within the BOD. The assignment is done in SAP ERP. When using SAP Extended Warehouse Management (SAP EWM) at the subcontractor’s warehouse, there is no loss of information about inventory and goods movements for the Service Parts Planning process.

Demand Planning

Within demand planning, the demand history of the parts is captured and the forecast is done with the help of extended and sophisticated forecast models. Forecast services cover the whole product life cycle from phase-in planning to product interchangeability to phase-out planning.

The demand history forms the basis for further forecasting and planning processes. Historical demand data is usually captured from SAP Customer Relationship Management (SAP CRM) or SAP ERP. During this upload process, the data is reorganized according to certain events such as:

• A future BOD becomes valid for a product
  
  
• A supersession (product replacement rule) is created for a product
  
  
• Stocking or de-stocking decisions have changed for a product

The system aggregates the demand for the product in different time buckets (weeks, months, fiscal periods) along with the BOD for each stockholding and customer facing location. The system allows manual adjustments to the raw data (sales orders from SAP CRM or SAP ERP) and to the aggregated demand data. The aggregated demand history serves as input into the forecasting process. This process is divided into the following steps:

• Aggregation of demand along the BOD (as part of demand capture and management process)
  
  
• Calculation of forecast at all customer facing and parent stockholding locations by use of a suitable statistical model
  
  
• Disaggregation of statistical forecast of the entry location along the BOD down to the customer facing locations. This disaggregation is done in proportion to the statistical forecast of the respective child locations.
  
  
• Either the statistical forecast or the disaggregated forecast can now be used for inventory and supply planning. Usually the disaggregated forecast is used because it has a broader database that delivers better results in statistical models.

The forecast service contains a huge number of forecast models. The selection of the appropriate model is based on the analysis of the demand history types such as constant, seasonal, or sporadic demand with or without trend. Before each forecast run, you can evaluate and change the forecast model where required. Table 2 provides an overview of the forecast models for Service Parts Planning.

Forecast service also offers scenarios for the phase-in and phase-out of a product. The forecast calculation for new and end-of-life products refers to the demand history of reference products which the user can define within a product group.

Inventory Planning

Inventory planning deals with the determination of economic order quantities (EOQ) in combination with safety stock. It also makes decisions about stocking or de-stocking of products and helps to identify surplus quantities in the network.

The inventory planning process determines the optimal stock and the optimal order quantity of a service part for a certain location. The overall aim is to ensure optimal service levels to the customer while keeping the inventory and purchasing costs as low as possible. During this process:

• Stocking and de-stocking decisions are made
• The EOQ in combination with safety stock is calculated
• Surplus and obsolete quantities are identified

The decision whether a service part is stocked or not stocked usually depends on the demand and the procurement costs. A simple example for a decision rule for stocking and de-stocking is shown in Figure 5.

Figure 5

Example for a stocking and de-stocking decision

The concept of a combined determination of an EOQ and safety stock helps to optimize inventory and procurement costs plus the targeted service level. The procedure always starts with the calculation of the EOQ. It is obvious that an increasing order quantity reduces the number of purchase orders, which also means a reduction in ordering costs. At the same time, the average inventory and the related costs increase. The EOQ calculation determines the order quantity with the minimal total costs (ordering + inventory costs) as shown in Figure 6.

Figure 6

Determination of EOQ

The amount of safety stock is dependent on the defined service level and the reliability of the forecast. The higher the forecast error, the higher the safety stock must be, to ensure the defined product availability. Other influencing factors are the replenishment lead time and the order quantity. Long lead times and small order quantities require a higher safety stock. Service Parts Planning uses sophisticated statistical methods to determine the required safety stock to supply within a defined service level. Once determined, EOQ and safety stock are used in the replenishment planning to calculate the procurement requirements.

Another feature of the inventory planning service is the identification of surplus quantities in the distribution network. The stock in a location plus all stock in transit for a service part within its BOD that exceeds the probable demand for a certain time frame are called surplus stock. Surplus stock can be removed (scrapped) by an interactive approval process.

Supply Planning

The results from the demand and inventory planning processes serve as input into supply planning. Supply planning determines the distribution requirements as procurement proposals. After the approval process, the proposals are handed over to the execution side.

Usually a distribution requirements planning (DRP) run occurs next — taking all demands and available supply as an input and creating as output the planned supply for the demands that are not covered. DRP determines the required replenishment quantities considering the output from the demand and inventory planning process. It is a multi-staged procedure that ultimately generates procurement proposals for the entry locations of the BOD. To determine the proposals, a net demand calculation is performed for all stock holding locations on each level of the BOD. This net demand is then shifted to the next higher level with consideration of the relevant lead times (e.g., goods issue time, transportation duration, and goods receipt time). To get a balanced planning, DRP also generates stock transfer proposals to cover the net demand at the child locations. Figure 7 is a simplified view of the net demand calculation process.

Figure 7

Simplified calculation of the net demand

Goods receipts at the customer facing locations are stock transfer orders coming from their parent locations. The same applies to the receipts at the mid-level of the BOD. These receipts are represented by stock transfer orders coming from the entry location. Both the purchase orders and the delivery schedules make receipts for the entry location.

The main demand driver at the customer facing locations is the forecast. Customer orders are only relevant as demand elements when they are dated in the future. Demand for the mid-level location is mainly represented by stock transfer orders plus stock transfer order proposals to its child locations. These planned stock transfers should cover the net demand at the child locations. Demand for the entry location is also generated by stock transfer orders plus stock transfer order proposals to its child location. These proposals satisfy the net demand at the mid-level. Finally, the net demand for the entry location is covered by procurement proposals (schedule lines or purchase requisitions).

The net demand at each level is always rounded according to EOQ, additionally considering other factors such as pack size. At the end of the DRP process, the responsible planner needs to approve the procurement proposals and hand them over to the execution side. From there, the proposals are transferred into purchase orders or delivery schedules for the scheduling agreement. Stock transfer proposals that were generated during the DRP run are not released to the execution system. The latter are recalculated in the subsequent deployment process and afterwards approved and executed.

DRP also supports the concept of consolidated ordering for virtual locations. Demands and inventories of the individual locations are aggregated and a consolidated net demand is determined for the virtual location. Supplier shutdowns can also be integrated in the DRP.

Distribution Planning

Finally, distribution planning decides the deployment of the received inventory along the distribution network. Inventory balancing is used when receiving does not cover the requirements in customer facing locations. The result of the planning process should guarantee that customer facing locations have enough inventory to fulfill customer requirements.

The purpose of the deployment process is the distribution of goods within the BOD (Figure 8). It is always performed from one level of the BOD to the level below. For deployment, you have to ensure that a demand is available on the child location and a distributable quantity is available for the parent location. If the BOD contains more than two levels, the BOD is split into sub-BODs and several deployment runs are performed.

Figure 8

Required deployment runs for a BOD

There are two ways of doing deployment, push and pull. When a service part is flagged for push deployment, the deployment is triggered by a goods receipt at the parent location. That means that the goods are not stored at the parent location but are immediately sent to the child location. This leads to a faster replenishment within the distribution network and it is often applied to fast movers. Pull deployment is triggered by a demand at the child location. This is the most common way deployment runs are performed.

The results of the deployment process are stock transfer proposals from all parent locations to their child locations. A key differentiator between the deployment in Service Parts Planning and supply network planning for sales products in SAP Advanced Planning & Optimization (SAP APO) is that the deployment process creates no stock transfer proposals for futures dates. This means that stock transfers are always due today.

The way they system determines the stock transfer is as follows:

• The available quantity for distribution is calculated for the parent location
  
  
• The net demand for the child locations during their replenishment lead time is determined
  
  
• If the net demand (during the replenishment lead time) is lower than the available quantity at the parent location, the system performs a fair share distribution for the remaining deployment quantity. Fair share distribution means that the remaining quantity for deployment at the parent location is distributed in proportion to the demands of the child locations
  
  
• If the net demand (during the replenishment lead time) is higher than the available quantity, the demand is divided into prioritized tiers (according to the type of demand) and sequenced. This ensures that demand with high priority is covered first. Within a tier bucket, the available quantity is also distributed according to a fair share-distribution.

The deployment process offers flexibility. If customer demand often differs from the actual forecast, it could make sense to retain a portion of the deployable quantity at the parent location. This allows the planner to react short-term to changing demand situations in the child locations. The planner could constrain the time horizon for the consideration of confirmed expected receipts or limit the supply quantity to the child locations. You can also trigger express (expedited) shipments, for example, to cover a demand in time through a faster but more expensive means of transportation.

Balancing Excess and Needs

Whereas deployment is distributing the goods along the defined BOD, inventory balancing can bypass the existing paths if you cannot fulfill a location’s need in time via the regular transportation lanes. This planning step can follow the deployment process in case the latter fails to satisfy all demands.

Inventory balancing can only be performed within a defined inventory balancing area which contains those locations that are allowed to supply via the predefined bypasses. Usually the shortage and the excess of a service part are calculated for all locations within the inventory balancing area and the imbalance is adjusted if possible. However, each adjustment is preceded by a cost and benefit analysis for the exceptional stock transfer.

Interchangeability of Products

The replacement of a service part by another is called supersession. Service Parts Planning supports (1:1), (1: many) and (many:1) supersession. These supersession types can be enhanced to complex replacement strategies using conjunction (AND) or disjunction (OR). Table 3 shows examples for replacement strategies.

Supersession in the Service Parts Planning calculates the key dates for planning with supersession. The process starts with the input of the pending obsolescence date (POD). It represents the earliest date the successor product can be used. Starting from the POD, the system determines the date when the remaining stock of the predecessor is used up (stock exhausting date). Doing a backward calculation starting from the stock exhausting date, supersession service determines the date when the planning for the successor product has to start. When the latter is reached, the demand history of the predecessor product is copied to the demand history of the successor product. This is then the basis for forecasting, EOQ, and safety stock calculation.

In an ideal situation, the stock of the predecessor product is sufficient to cover its demand and the supply for the successor product is already available to satisfy the requirements for the successor product. If an imbalance occurs, DRP can generate substitution orders to cover the successor product demand with supply of the predecessor and vice versa. If a substitution strategy exists for remanufactured products, DRP also creates substitution orders (already for the new product that has not yet been used) to solve the mismatch between the demand and remanufacturing capacity on the supplier side.

Based on the fact that Service Parts Planning is an exception-driven process (due to the high number of stock keeping units [SKUs]) and that planning runs are mainly performed in the background, process monitoring and analysis are extremely important. Monitoring and controlling are done with the help of shortage and alert monitors that report problems or imbalances in the supply chain. Shortage and alert monitors are part of Supply Network Collaboration (SAP SNC, formerly known as SAP Inventory Collaboration Hub [SAP ICH]) introduced with SAP SCM 5.1. Performance measures for the customer within the supply chain and from the supplier to the manufacturer are included in SAP NetWeaver Business Warehouse (SAP NetWeaver BW, formerly SAP NetWeaver Business Intelligence [SAP NetWeaver BI]).

Ute Messmer

Ute Messmer studied industrial engineering with focus on information technology and operations research at the University of Karlsruhe, Germany. Before joining Westernacher and becoming a consultant, she worked for Hewlett-Packard, Germany, in several supply chain operation, design, and management functions.

You may contact the author at spm@westernacher.com.

If you have comments about this article or publication, or would like to submit an article idea, please contact the editor.