SAP IBP 1908 brings some much awaited updates to the advanced supply chain planning platform. Some are marginal enhancements while others are significant changes to the tool and is a welcome relief to many IBP users.
What is new in IBP 1908? Here is a summary of the changes to the platform.
1908 update covers multiple touch-points: Inventory Optimization, Supply Planning, order-based planning, Demand Planning, S&OP module, DDMRP and SAC – the SAP Analytics Cloud.
The changes could be broadly classified as below:
Sales and Operations Planning
1908 update covers an approach oriented more towards planning by exceptions by allowing the user to draw reference lines that can be either average, maximum or minimum of a key figure. Although this is a welcome addition, our recommendations in such cases will be towards alerts based on our diagnostics. This allows users to clearly see exceptions.
By adding the maximum and minimum, it enables the threshold to be dynamic.
IBP 1908 allows users to create additional local what-if type scenarios. You could add them in the web interface or Excel add-in. This allows you to re-run the scenario without affecting the operational data in the tool.
Here is a much-awaited change to the Demand Module.
After repeated requests to SAP product managers, SAP is now introducing the Seasonal Linear Regression model into the Demand Module.
This is a powerful yet simple model to use for time series with stable patterns on both trend and seasonality.
Here is the APO version of the model from the SCM product. Hopefully, the results from IBP 1908 enhancement matches the output of the old APO system.
SLR works well and is low cost to use largely because it does not take a lot to initialize the model.
Good move SAP. However there are still some small issues – IBP has defined user inputs for the phi coefficient here; interesting see a smoothing coefficient in a linear model …. neutral mutation?!
Consider Time-series Properties
In APO, DP Auto Model digests the demand components and comes back with messages that tells a story about the model selection. Some of the typical message used to read “Auto Model has selected a Trend Model”
The new addition of checkbox for considering time-series properties is allowing the engine to exclude certain models and focus on time series components. The typical time series components –
- Level – to gauge the on an average sales
- Trend –to see the decline and growth
- Seasonality – to judge peaks and valleys afterwards
- Random noise and then the order (or shipment) frequency components for intermittency
However, this seems like a method for excluding specific models and may cause some surprises in reality.
SAC Integration into IBP
IBP offers integration into financial planning through the SAP analytics cloud. You can integrate the supply chain planning data from IBP and financial data from S4. Similarly, Budgets and plans can be brought over through this interface. This should allow a true 360 view of the business and facilitates a complete IBP process.
This will truly make IBP successful with finance view in the dashboards. But the use cases will need to be further developed for the trigger points for the demand planners and finance executives.
IBP covers the Demand driven replenishment process to enable tactical planning and replenishment. Average daily usage is now calculated based on data from ECC and IBP and buffers are created that use both order quantity as well as forecasted demand in the future. A new Demand Driven Replenishment operator is added to the IBP menu structure.
Now supply planning enables allocation of resources using a capacity consumption rate as the conversion factor.
New Merger policy Key figure is introduced that can decide whether to use Downstream or upstream values to predominate in the supply synchronization.
Inventory optimization is better enabled through the introduction of demand propagation. This now introduces two new sources – transportation propagated demand and BOM propagated demand.
You can refer to additional changes at the SAP doc available here
If you would like to discuss IBP implementation or optimization of the algorithms, please feel free to reach us.