First delivery of live embedded SAC analytics in SAP Predictive Asset Insights 2011
Introductory note: just in case you are wondering about the product mentioned in the title; the product previously known as SAP Predictive Maintenance and Service (PdMS) as of the 2011 release now includes the simulation capabilities of SAP Predictive Engineering Insights enabled by Ansys, and was renamed to SAP Predictive Asset Insights (PAI).
A new step in PAI analytics
With SAP Predictive Asset Insights 2011, just recently released, we are going the second step in our journey to embed analytics screens provided by SAP Analytics Cloud (SAC) into PAI business screens.
PdMS 1902 – Offline Analytics
The first step was in PdMS 1902, when we offered configuration capabilities to show SAC embedded in five screens:
- A new Analytics Dashboards application to show dashboards across a fleet of equipment
- New sections on the master data pages of equipment, models, locations, and spare parts, to show a dashboard with data for just that business object the user views.
All screens allow to show SAC dashboards (or “stories”, as they are called in SAC) in view mode, without any limitation to the functions afforded in SAC. Below see an example from the equipment application. All data is automatically filtered for the equipment, here this Pump 554. SAC occupies a part of the screen. Below the Analytics Dashboards section with its SAC story starts the next section (unrelated to SAC).
In PdMS 1902 analytics were limited to “Offline Analytics”: the customer had to copy business data from PdMS to their SAC tenant, using one or more of several public PdMS OData APIs. The customer could set up a schedule in SAC to have the data refreshed, but the analytics would not really be live. If during an hourly refresh cycle a new equipment would be created, or two new notifications, the analytics dashboard would not immediately reflect this.
For the SAC setup the customer must have its own SAC enterprise license, set up a connection to a PdMS/PAI system, create SAC models, and SAC stories (the dashboards).
PAI 2011 – Custom Live Analytics
Cue to PAI 2011 which brings several improvements:
- New data sources – PAI now allows to show alerts and sensor (indicator) time series data
- Aggregations of numeric sensor data – as raw sensor data (e.g. collected every second) can be far too big to use in analytics, any numeric values get aggregated into hourly buckets. PAI calculates the minimum value, the maximum, the average, the sum, and the count of values. Raw sensor data is not available for now.
- (Almost) Live analytics – apart from the hourly aggregations any other data shows in analytics a very short time after it was created in PAI
In PAI 2011 we introduce a flavor of live analytics we call “Custom Live Analytics”. This is for all customers to whom it is not enough to show the pure PAI data in analytics, but who want to augment the data with their own data (e.g. shop floor production data related to the equipment maintained in PAI, to see what the equipment was used for), and who want to run calculations on the PAI data (e.g. to calculate a special health index from the equipment sensor data and any health data from PAI machine learning engine algorithms).
For Custom Live Analytics PAI affords to access all its data in a consolidated database, and customers use their own HANA system to access the data and do their custom augmentations.
The rest of the analytics story is the same as in Offline Analytics: the customer creates their own connections, models, and stories, and shows the stories in PAI.
SAP Help is a good source for reading more about this
- Overview on analytics in PAI
- For the technically inclined: how to configure all this
A later PAI release – Standard Live Analytics
We plan to add another larger new capability to PAI, which would be for customers to whom it is not important to augment the analytics data, who just want to use what they already have in PAI. The customer can use out-of-the-box data views but cannot change them. The customer can use standard analytics functions with the rich set of charts in SAC, drilldowns, linked analyses, etc., but cannot access more advanced SAC functionality like forecasting. The benefit to the customer is two-fold:
- Apart from creating their own SAC dashboard, the whole data infrastructure is ready-to-consume
- The customer does not need to buy an SAC enterprise license anymore
We call this next step “Standard Live Analytics”.
We have ideas for a range of additional features, such as being able to configure which dashboard to show dependent on the equipment type, or to allow “click-throughs” from a business object in analytics (like a bar in a bar chart representing one equipment) to the to the business application screen, or to afford the last indicator value, or to provision a small set of raw indicator data on demand, to analyze e.g. the data around a specific alert.
What are your desires?
In the business area where analytics of traditional asset management data (equipment, notifications, work orders, failure modes, …), plus sensor data, plus alert data overlap, what are your needs to show charts and trends? Which business processes could you make better with integrated analytics across these three sets of data?
Let me know … I am happy to listen!
My desire when it comes SAP IAM, I would like to establish appropriate linkage of Failure modes to Maintenance strategy…with this I feel, maintenance team can really evaluate effectiveness of current maintenance strategies and also they can easily identify the GAP…
I feel with this, when we do "Check List Assessment/FMEA/RCA" , they can address the GAP in easy way and formalize the recommendations with PM program…
glad to hear that you see value in the SAP IAM functions. And yes, focus on SAP Asset Strategy and Performance Management (ASPM) when it comes to looking at failures and failure modes and how they should impact the maintenance strategy.
Also check out our new failure curve analytics in SAP PAI (fka SAP PdMS) at https://blogs.sap.com/2020/11/17/new-weibull-based-pof-curves-in-sap-predictive-asset-insights/. We are still looking at how to make use of its output in ASPM.
The help for "Live embedded SAP Analytics Cloud (SAC) analytics" mentions that this feature is only available on AWS. Does this mean that SAC needs to run on AWS (instead of Neo), or PAI, or both? If it is PAI is feature parity between Azure landscape and AWS landscape planned?
yes, SAC needs to be on AWS, and it is my understanding that today SAC does not exist on Azure. We are looking at ways to bring PAI analytics end-to-end onto Azure as well. PAI itself runs on AWS or Azure.