Understanding Key Differentiators of SAP IBP for inventory

As part of the key enabler # 1, end-to-end supply chain planning in one tool, SAP IBP for inventory provides you with several approaches. Here are three different possibilities:

Integrate previous processes supported by SAP IBP, for instance:

• Use ABC-XYZ Segmentation for product portfolio classification.
• Use the Demand Planning Process to establish a global demand plan.
• Connect Demand Sensing to estimate demand forecasts in days updating values in the upcoming 0–8 weeks.

In addition, exploit the power of SAP IBP for inventory through the recommended steps for an inventory optimization process according to the SAP best practices for supply chain planning:

1. Validate data inputs: Use features and functionalities to validate inputs for inventory optimization, for example, demand forecast, coefficient of variation, batch sizes, lead times, and so on.
2. Run and review an inventory plan: Activate and plan, in batch mode or interactively, an inventory planning operator, for example, a multi-stage echelon or a single-stage echelon optimization, and calculate the inventory target components. Further, use the Application Jobs app to schedule the inventory optimization according your business requirements.
3. Run a what-if scenario analysis preceding the inventory plan publication: Before approving results involving other functional areas, recreate possible future situations ad hoc via simulations. Recalculate all key figures and save the values through scenarios, or redefine the structural characteristics of the planning model via Versions.
4. Finalize your inventory plan connecting to other functional areas as marketing, sales, operations, procurement, and so on. Communicate results locally, regionally, or globally along all functional areas through enhancements of the planning views within the SAP IBP add-in for Microsoft Excel, or via analytics and dashboards.

What is coming up next?

• Supply Review – Heuristic and Optimizer: Use the outputs from Inventory Optimization to calculate and revise a constrained demand plan via the heuristic and the optimizer run.
• Or, integrate an external processes, for example, operational planning and execution processes.

We need to remember that location-centric inventory planning oversimplifies the situation leading to:

• Over-buffers of inventory.
• Bullwhip effect.
• Limitation to determine any postponement strategy.
• Not being able to handle BOM, lot sizes, and other supply chain complexities.

SAP IBP for inventory simultaneously optimizes inventory across the end-to-end supply chain, as follows:

• Coordinated planning eliminates inventory over-buffering while meeting service-level objectives.
• Internal service-level optimization provides significant inventory reduction.
• Demand variability propagation to upstream stages avoids bullwhip effect.
• Streamlines centralized inventory planning.

We need to understand that when we are talking about inventory, we are also talking about the inventory location and inventory form as :

• Raw materials inventory
• In-process inventory
• Postponement inventory
• Distribution inventory
• A sum of all these inventory values

On the other hand, when we are talking about inventory, we are also touching the purposes for what it's being held, for example, safety, cycle, pipeline, pre-build, minimum required, and total stock. All of these inventory types are influenced by drivers as follows:

• Safety stock: Demand, Demand uncertainty, Lead times, Lead times uncertainty, Review frequency, Service Levels targets and Service times.
• Cycle stock: Demand, Review frequency, Batch sizes and Production rules.
• Pipeline stock: Order processing lead times, Transit times and Demand.
• Pre-build stock (Enabled by the module SAP IBP for response and supply. A module that we will not address in detail in this training.):
  • Time-varying capacity
  • Time-varying demand
  • Sourcing ratios
  • Changes in safety stock
• Minimum required stock: business drivers.
• The sum of all these values: the sum of all drivers.

SAP IBP provides a three layer approach to execute changes from integrated data sources, such as an ECC:

• Version
• Simulation
• Scenario

The following table simplifies the characteristics of Versions, Simulations and Scenarios:

If we compare the traditional approach to calculate the inventory target against the state of the art, we come to the following conclusions:

• The real world is stochastic, that is, variable and unpredictable. It is not deterministic.
• The real world is multistage, and it does not work with isolated single stages.
• In the real world, time varies, and we cannot assume that time is stationary.
• The real world needs a comprehensive data model, moving far away from multiple or limited data models.
• The real world implies an enterprise scale, and not only a desktop scale.
• The real world needs a dynamic integration, not a static offline analysis.

Artifical Inteligence (AI) is changing the world how business are build. Possibilities not even imaginable a few years ago become available for nearly everybody.

SAP Integrated Business Planning offers strong capabilities in the environment of mathematical optimization and machine learning functions:

In the future, an even deeper integration is planned with the inclusion of Large Language Models (LLMs), which will change the way how users will approach planning topics within systems.

Looking at the detailed AI-based functions within SAP IBP, you notice that various of these are cross-application options while on top application-specific AI functionality is available as well:

In the area of monitoring, AI can be applied for job anomaly detection and in the automatic detection of alert thresholds. Related to master data, there is a function available that recognizes patterns in the master data and gives recommendations for the entries which are detected as not fitting to the patterns.

Probabilistic planning is applied to handle uncertainties in planning. Particularly in inventory planning two major kinds of uncertainties need to be taken into consideration:

• Demand Uncertainty: this uncertainty reflects the unknown size of future demands (e.g. customer order quantities).
• Supply Uncertainty: uncertainty on supply side is related to e.g. the variability of lead times.

IBP for Inventory takes these uncertainty into account during the optimization of the inventory plan (Multi-stage Inventory Optimization). On top of this, it has AI-related capabilities to recommend lead times including its variability based on historical observations (the demand uncertainty-related analysis results can be taken over from IBP for Demand).

In inventory planning planners are facing the challenge to find out how adjustments to safety stock values affect service levels, for example, if you want to restrict your investment in inventory levels at the end of a fiscal period. The Service Level Prediction operator allows planners to predict the customer service level by product location based on a predefined safety stock plan.

After creating an inventory plan, planners can analyze predicted customer service levels in the future periods of the plan. Measures can then be taken to minimize the impact during the critical periods of inadequate service to protect customer service levels and minimize impact on revenues.

Other examples, in which these kind of functions can be used, are unplanned events that increase lead time or reduce capacity, potential new market demands or unexpected demand variability.

On top of what is offered in SAP IBP for Inventory directly, various functions in the cross-application part of SAP IBP are AI-enabled as well, as the following list of examples shows:

• Segmentation: Segmentation, a.k.a. classification, helps to segment/classify objects like product/location combinations according to their characteristics. The goal is to have homogeneous segments/classes with similar characteristics, that can be treated in planning in a similar way. Common examples for this kind of approaches are ABC classification (according to importance) and XYZ classification (according to variability). SAP IBP enables the usage of the AI algorithm called k-means, which segments the objects into classes as homogeneous as possible.
• Master data consistency check: Master data quality for planning can be improved by self-learning semantic rules based on the identification of problems and the recommendation of correction. This function can be executed as an application job using a dedicated job template (ML Master Data Consistency).
• Alert threshold determination: The AI approach used for segmentation can be applied in exception-based planning by using the same kinds of algorithms for the determination of alert thresholds. Beside the already mentioned machine learning algorithm k-means also DBSCAN can be used.
• Batch job anomaly detection: This AI-based function detects and reports anomalies in job runs.
• Re-allocation of hardware (Cloud operations): As cloud-based solution SAP IBP can benefit from flexible hardware re-allocation. Artificial intelligence evaluates past job runs and automatically dispatches new jobs to the environment with the best fit, including consideration of the needed size of the elastic computer node. This approach leads to better performance due to taylored hardware sizes with more parallelization and lower costs, as the hardware capacity increase only happens when needed.