S/4HANA and Embedded Analytics is steadily gaining shape in the market.
Predictive capabilities have been incorporated in the S/4HANA system since the cloud version 1708
In in this blog, am trying to explain these concepts, with the help of a 1709 on Premise system.
If you have the relevant roles, you can add the specific tile from the APP Finder in Fiori Launch Pad.
Go to App Finder and look for “Predictive Models- Manage predictive models for business applications” App, which can further be added to the required group.
Click on the APP to see more details.
Unfortunately, am facing some issue here.
Let’s try to solve the error first.
Go to Tcode /n/iwfnd/maint_service in the Front-end Server and add the service “ANA_PAI_REPOSITORY_SRV”
Provide the corresponding System Alias name and click on "Get Services"
Search for your service name.
Add the selected service.
Provide the valid Transport connection details. Here am giving only $TMP
Now you can see that the service is activated successfully.
Check if you are getting Status code as 200.
Once the service was activated successfully, I am now able to see the standard content.
There are some use cases like “Operational Contract Management” and “Stock Management” which has been introduced in 1709.
SAP ship these Predictive Scenarios as standard with SAP S/4HANA 1709. They don’t ship trained models but a template that enables the customers to train models on their own data.
Predictive Model 1:
Predictive Analytics for Stock in Transit in Warehouse Management
With this predictive app, customers can track materials in transit/open stock transport orders for which no goods receipt has been posted by the receiving plant yet which have already exceeded the estimated time in transit.
The predictive model training needs to be done by Analytics Specialist. The application Overdue Materials - Stock in Transit needs the role of Inventory Manager.
In this blog, I have taken you through the series of steps to see the Predictive models in the S/4HANA system. In my next blog, I will try to leverage these standards and test/train some sample custom data.
For more details about the related 2 apps(especially on the Contract consumption), I would Strongly recommend you to go through the following blogs aswell.
Some FAQ on this topic:
a) Will customers get Pre-delivered scenarios(mentioned above), by default if they have a S/4HANA license. Or is there any specific license required for the same?
Remarks: Pre-delivered scenarios including the ones mentioned above are part of S4HANA and they would work without additional licenses of PA or PAi.
b) Apart from the above ones, will customer have the flexibility of adding a new custom PA scenario utilizing any of the broad algorithms like regression, segmentation or so on?
Remarks: For that customer needs PAi . Specifically, PA Application Edition license today. If they want to adjust existing predictive models or want to add new scenarios with new models – they would need this additional PA license. They would first create models in PA like we do today and bring to S4HANA using PAi functionality. Although from PAi side, segmentation is not supported yet. But on roadmap.
Since S/4HANA 1709 EHP1 On premise customers can purchase a PA App Edition License and build new Predictive Scenarios in PA 3.3 and publish them in to PAi in S/4 as a new predictive scenario.
c) I understand that there is a PAI (Predictive Analytics Integrator) engine within S/4HANA core. Will we have this Engine available for all S/4HANA licenses.
Remarks: There is PAi engine technically to have the pre-delivered predictive scenarios in S4HANA: NetWeaver ABAP and HANA DU components. Currently, there is no restriction on S4/HANA licenses .
Again the same thing is if customers want to change /adjust/ build new models then you would need PA Application Edition license which will include PA + PAi.
In fact Pai + APL are shipped with S/4HANA as standard the PA App Edition License gives you access to the PA suite to build new or edit existing predictive scenarios.