The supply planning process creates a feasible forward-looking production plan on a weekly basis by considering the capacity constraints, material constraints, manufacturing run rules, and supply chain network. A typical supply planning process is executed every week in weekly buckets for a six- to 12-month horizon, but may differ from company to company.

Having an accurate supply plan can help businesses with increased visibility to the entire supply chain, shorten order-fulfillment times, reduce inventory levels, and improve customer service. Business requirements in the process industry need to be adequately addressed by the planning engines. Some of the critical challenges when using Supply Network Planning (SNP) enabled by SAP Advanced Planning and Optimization (SAP APO) Capable to Match (CTM) as the engine to generate a weekly supply plan are:

• Accounting for lot size and available capacity in a weekly bucket to maximize capacity use
• Prioritizing demand based on cumulative lead time (time required to move the product from the manufacturing location to the distribution center)
• Minimizing manual intervention when generating a supply plan and performing an inventory build for seasonal products

We first discuss the critical business considerations and questions that need to be answered with respect to supply planning in the process industry. We also describe how with the help of innovative changes to the existing functionalities combined with enhancements to planning engines using various user exits and Business Add-Ins (BAdIs) available in APO CTM, we were able to address these business challenges. We cover how to better solve business challenges using CTM as the preferred engine.

Introduction to SNP and CTM

CTM is a planning engine that enables multi-level supply and demand matching as part of SAP APO. CTM offers two modes containing functionality for use in Production Planning and Detailed Scheduling (PP/DS) and SNP.

For the purposes of this analysis, our focus is on the SNP CTM solution. CTM is one of the three major algorithm options companies have when they are implementing SNP for mid- and long-term planning. The basic characteristics of SNP CTM include the following:

• Order-based planning: Every demand that is stored in the system as a liveCache order is planned individually by the CTM algorithm. The algorithm provides the capability to generate logs that assess each forecast, sales order, or other demand source individually as being met on time, met late, or not met at all. This differs from the time-based, bucket-oriented planning that is used by SNP heuristics and the SNP Optimizer.
• Demand prioritization: Offers the capability to establish location-based or customer-based demand priorities that drive the solver to consider certain demands in priority sequence during the CTM planning run
• Finite capacity planning: CTM generates a constrained supply plan that can consider transit times, goods receipt/issue processing times, transportation capacity, and production capacity for in-house manufacturing
• Supply prioritization and multi-sourcing: When multiple sources of supply are available (e.g., multiple lines at the same plant, multiple in-house manufacturing plants, or a make-versus-buy scenario) CTM can consider a prioritization sequence before scheduling the supply to meet the demand. In scenarios in which capacity is constrained, CTM defaults to planning a receipt in advance before it seeks out a lower priority supply source. The impacts of this phenomenon are discussed in the “Seasonal Pre-Builds” section.

Overall, CTM is becoming a more popular choice for companies in the process industries that seek a middle ground between the simplicity and limited capabilities of SNP heuristics and the additional master data requirements and solution complexity associated with successfully implementing the SNP Optimizer.

Key CTM Requirements in the Process Industries

Based on our experience at multiple companies implementing CTM, there are some commonly reoccurring requirements that span process industries. These requirements can pose challenges during CTM implementations if they are not properly considered at the onset of the implementation life cycle:

1. Capacity planning at a weekly level – frequently referred to as master scheduling in the process industries, mid-term planning is usually executed in weekly buckets starting two to six weeks out from the present date and spanning through the conclusion of a six- to 12-month horizon. Generally in a make-to-stock environment, only forecasted demands are available in that horizon. They are loaded into the supply planning engine at a weekly level of detail. Additionally, run rules at the manufacturing plants are used that result in the maintenance of prescribed minimum lot sizes at the product-manufacturing location level. The key requirement is to generate a master schedule that considers pooled capacity in a weekly bucket and plans forecasted demands, while also considering the minimum lot size at a weekly level for all products at their respective manufacturing locations.
2. Demand prioritization: Unlike some industries that respond to specific customer demands and orders, process industries generally do not consider customer-specific demand prioritization in the mid-term supply planning horizon. However, the supply planning solution must have the capability to sequence demands based on transit times (this is particularly important when supply networks are highly complex) and safety days’ supply. In process industries, safety days’ supply is used much more frequently than safety stock in units. Thus, the requirement to plan a certain number of days in advance for certain demands is very common. This includes the ability to maintain a location-specific quantity for safety days’ supply that may vary widely across distribution locations for a given product.
3. Seasonal pre-builds: Particularly in the consumer products industry (and to a large extent, the food and beverage sector), seasonal demand spikes give rise to unique requirements around pre-building inventory. The two key requirements relating to seasonality of demand are:

• Inventory build should level-load capacity consumption so that labor usage can remain constant and machine capacity requirements can remain stable (this prevents unnecessary capacity investments)
• Shelf life of the product should be respected when proposing a build-ahead to prevent wastage

Capacity Planning at a Weekly Level

When CTM is executed in SNP mode, the engine assumes that capacity availability in the resource master is considered in daily buckets. This is regardless of the demand or supply aggregation settings used in the CTM profile. Moreover, the planning engine does not generate multi-day planned orders. If insufficient capacity exists at a daily level to produce the minimum lot size for a planned order, then no planned order can be generated. Figure 1 shows how capacity planning works in standard CTM.

Figure 1

The standard capacity planning process in CTM

In Figure 1 notice that in Week X, the demand for that week can be met by the available capacity. However, in Week Y, demand is dropped by the CTM engine because the lot size settings are considered daily by default. At a rate of 1,000 cases per hour, the weekly bucketed capacity for the resource is 80,000 cases (16 hours, 5 days per week). However, because CTM cannot exceed the capacity within an individual day, capacity is consistently left “on the table” when lot size settings do not round evenly to daily capacity constraints. Note that this issue does not occur when planning for production at a daily level of detail.

While demand aggregation (e.g., to a weekly level) can help to group demands into a weekly bucket, it is not possible in standard CTM to prevent the under-utilization of capacity in this scenario. This condition exists (though on a smaller scale) even when minimum lot sizes and rounding values are decreased to be a smaller proportion of a single day’s capacity.

Using an enhancement, you can customize CTM to add together the available capacity in hours for an entire week (this capacity can be based on SAP standard weeks or fiscal weeks, as appropriate). In Figure 2, observe how the entire weekly capacity in hours is considered as continuously available capacity across the entire week. In Week Y, the engine is able to accommodate the additional demand requirements. For illustration purposes, the graphic shows the entire capacity as available on one single day within the week.

Figure 2

Capacity planning after an enhancement

Executing the CTM engine using this enhancement makes it comparable to how the SNP capacity leveling algorithm perceives resource capacity. However, it does so without sacrificing the many benefits to using CTM over a capacity leveling solution (including consideration of shelf life, multi-sourcing with using source of supply priorities, and restricting the lateness of receipts to meet demand – to name a few). Without this enhancement, the CTM engine in SNP mode generates a sub-optimal solution when production capacity constraints exist. This is because generally the combination of factors that cause the issue – minimum lot sizes, rounding values, with the ability to continuously manufacture – are almost universally present throughout the process industry.

The following change is required in CTM to accommodate weekly bucketed capacity planning: The CTM data selection process was enhanced using the BAdI /SAPAPO/CTM_RESOURCE with Method= RESOURCE_MODIFY.

This BAdI ensures that daily resource capacity is aggregated into a weekly bucket before CTM planning is executed. Note that if constrained capacity planning across horizons is done at different levels of detail (e.g., daily or weekly) using CTM, then you need to call the BAdI iteratively to aggregate resource capacity based on the CTM time stream.

Enhanced Demand Prioritization and Planned Production

In Figure 3, a simple supply chain is illustrated with two destination distribution centers and one source manufacturing location. The transit time to distribution location 1 of 14 days is 12 days longer than the transit time to distribution location 2. Additionally, the safety days’ settings vary by product location. At location 1, the setting is maintained as seven days, while it is 14 days at location 2. The minimum lot size at the manufacturing location is 2,000 cases, with demands in the future of 500 cases on Day 51 and Day 50, respectively.

Figure 3

Demand prioritization and planning production in standard CTM

After examining Figure 3, ask yourself “Are two planned orders needed to fulfill the demand on time for both distribution locations?” The correct answer to that question is no. However, standard CTM prioritizes demands based on date (considering all else is equal) without considering transit time as well as safety days’ supply from the product-location master. In this example, ideally only one planned manufacturing order would have been generated on Day 29. The demand on Day 50 at location 2 was prioritized ahead of demand on Day 51 at location 1. So going in sequence, the CTM engine generated an answer to the question of how to produce at the manufacturing location just in time to meet location 2 demand. The engine does not have the level of sophistication required to remove the unnecessary over-production after an additional (and sufficient) planned order is created on an earlier date that is capable of fulfilling both demands at both locations.

This poses several problems. First, over-production in supply planning can result in false-negative signals regarding capacity consumption. Second, if acted upon and not reviewed by a planner, increased inventory (an adverse balance sheet impact) and increased wastage can result. This example is illustrated with a simple scenario. As network complexity increases, this issue can lurk behind the scenes and cause much frustration when attempting to analyze a root cause for the system proposing over-production.

Using a combination of standard CTM profile settings and an enhancement, you can avoid these suboptimal planning results. In Figure 4, note that some additional criteria have been added in red to the demand priority determination. In standard CTM with default settings in place, only the demand date is used to sequence and prioritize demand.

Figure 4

Demand prioritization and planned production in enhanced CTM

Due to the resequencing of demand priorities, the CTM engine initially creates a planned order based on Demand 1. The Planned Order on Day 29 will also fulfill the requirements associated with Demand 2. In the previous scenario in Figure 3, the sequencing of demand priorities resulted in the creation of a planned order which was just in time for Demand 2 but too late (Day 34) to meet the necessary transit time and safety time required for Demand 1. The assumption above is still valid: The demands at both destination locations are equal in priority in all other respects aside from forecasted demand date (e.g., customer and location priorities are all equal).

These two changes are required to achieve the above planning run result:

1. Use a planning parameter: As explained above, standard CTM prioritization uses the demand due date as input, which in our example is Material Availability Date at the destination location. This results in a suboptimal supply plan. With parameter PRIOBYTOTALLEADTIME it is possible to modify the demand date by subtracting the total transportation duration for all transportation lanes of the supply network that end at the demand location. Then CTM prioritization is applied for the modified date. To define a Planning Parameter in a CTM profile, navigate to the Control Menu and select the option for Planning Parameters from the dropdown. The exact text must be manually entered into the Parameter Name field, followed by the text value of X in both the Value 1 and Value 2 fields (Figure 5). This indicates that the standard planning parameter should be in use when executing a CTM run using the CTM Profile currently selected in the system.

Figure 5

CTM planning parameter maintenance

2. Use an enhancement to demand prioritization: You can use the above parameter to calculate the ship date based on the total lead time, but it does not account for safety days. An enhancement is required to consider the safety days’ supply field maintained in the product location master when sequencing demands. This is needed only if safety days’ are used in planning (both in the CTM profile and the product master). However, this enhancement is required to prevent over-production to work in concert with the transit time planning parameter for demand prioritization.

The default CTM sort profile was modified using user exit APOBO020. With this enhancement we were able to add safety days’ supply to the overall demand date calculation. In Figure 6, the sort profile is updated to consider the user exit criteria.

Figure 6

Sort profile utilizing user exit priority

The equation we used for the demand date for purposes of prioritizing demands in the CTM planning run is as follows:

DD_S = MAD_D - TLT_N - TSD_N

Seasonal Pre-Builds

Thanksgiving dinners, Oktoberfest beer festivals, and ski season knee injuries are all examples of annualized cultural phenomena that drive seasonal demand. It is difficult to find companies in the process industries that don’t struggle to generate mid-term and long-term production plans for seasonal products without significant manual intervention. Using SNP CTM, supply planners can manage the master data inputs rather than re-write the transactional outputs to generate schedules that level-load capacity across the entire year while respecting shelf-life considerations.

The primary master data lever that can be used to manage the requirements when using finite capacity planning in CTM is the order creation frame (OCF). OCF is defined as the duration of time (in days) in advance of the demand date that is permitted to generate corresponding orders in the system to fulfill supply. If a demand is on day 50, for example, a 28-day OCF means that Day 22 would be the earliest day on the horizon that CTM would consider creating planned production to fulfill the Day 50 demand. It is worth noting that this OCF applies across network levels and bill of materials (BOM) levels, so it must be inclusive of transit time and component lead times (as can be observed in Figure 7). Additionally, by default, CTM first attempts to fulfill demand as close to just in time as possible. In a seasonal pre-build with capacity constraints, however, the boundary for allowable earliness must be carefully considered.

Figure 7

How CTM considers the OCF

 The OCF is distinguishable from the method by which CTM considers shelf-life settings in the product master. In the CTM product master, shelf-life settings are narrowly focused only on the location level and do not consider transit times and multi-tier network dependencies. If OCF is left blank, the system defaults to the global setting (which we recommend to be 999 to prevent its being restrictive). As noted in Figure 8, the OCF can be maintained by navigating to the product location master (via transaction code /SAPAPO/MAT1 – Product Master) and selecting the SNP 2 tab (CTM Settings sub-section).

Figure 8

The OCF maintained in the SNP product master

1. It must be short enough to prevent overproduction. Figure 9 illustrates that in a scenario in which multiple possible sources of supply exist (i.e., multiple production lines), it is possible to define an OCF that is too long. This results in the underutilization of capacity on resources that are lower in the priority sequence. It also builds up unnecessary inventory in advance, which will have an adverse balance sheet impact for working capital.
2. It must be long enough to prevent under-fulfillment of demand. If planning book alerts and CTM logs indicate that shortages are resulting from the CTM planning run, then the OCF, to allow for seasonal builds, may need to be increased to allow for more pull-ahead (Figure 10).

Figure 9

An OCF that is too long

Figure 10

An OCF that is too short

Supply planners should assess the above boundaries and arrive at the right answer to generate an optimal plan that minimizes excess inventory while preserving demand fulfillment. Analyzing historical demand and production levels is a good place to start with this analysis. However, short of a full-scale data analysis, our recommendation for practical implementation is to start with a higher value based on the shelf-life properties and BOM structure of the product, then monitor projected inventory and capacity utilization on alternate lines to reduce the value over time to find the optimal setting.

The following master data and CTM settings are also relevant to facilitating a planning run for seasonal pre-builds:

• Navigation to the ECC Material Master: Transaction code MM01/02/03, Plant Data / Stor. 1 Tab
• Navigation to the SAP APO Product Master: Transaction code /SAPAPO/MAT1, Properties Tab

• Navigation to the CTM Profile: Transaction code /SAPAPO/CTM, Pegging Type setting
• Navigation to the SAP APO Product Master: Transaction code /SAPAPO/MAT1, Demand Tab, Pegging subsection

• Navigation to APO Resource Master: Transaction /SAPAPO/RES01, Capacity Profile

As we have discussed throughout this analysis, the emerging prevalence of CTM across the process industries results in the need to carefully analyze requirements and desired outcomes at a low level of detail. The best implementation practices for a supply planning solution focus not only on designing a system that meets the business requirements but also – given the complexities of the planning processes in SAP – focusing on deriving what those requirements should be.

When companies can trust a planning engine to generate an output that reliably prevents overproduction while satisfying future customer forecasts, then user adoption of the system is maximized and manual correction errors are minimized. Once the tactical and focused topics of weekly capacity planning, demand prioritization, and seasonal pre-builds are examined through the lens of planning in CTM, they are essential to preserving and maximizing the business benefits – both income statement and balance sheet – associated with a supply chain transformation.