Today, I want to discuss some of the differentiating elements and dimensions in our approach to predictive maintenance. Of course this topic is hyped everywhere and many vendors offer solutions for maintenance and service. Nevertheless, SAP’s solution, based on SAP HANA, clearly provides three main technological advantages.

1) Build on top of one development platform being able to process huge volumes of data in real time and offering several capabilities (beyond a pure database)which are essential for predictive maintenance, e.g. predictive analysis, text search and geospatial data

2) Integration with existing back-end business processes: follow-up activities derived from predictive analysis insights are triggered directly in back-end (e.g. scheduling maintenance)

3) Individual implementation of dedicated, product or assets related predictive algorithms and highly intuitive user interfaces for dedicated roles without any need for compromise.

Predictive maintenance and service integrates different data sources, from business systems like SAP ERP or SAP CRM, with operational data capturing sensor information directly from the machines. As you can imagine, huge amounts of data must be processed in real-time to ensure accurate correlations and analysis along dimensions you have never looked at and thought of before. SAP HANA with its in-memory technology, not processing data on any kind of  aggregate, ensures that all different data sources are brought together in one instance and transparency in any direction is made possible.

Having the data together in one data model is a first step; the power of predictive analysis on top is clearly the second and more relevant as well as challenging one. SAP HANA includes a prediction engine, with a library for predictive analysis, giving the data scientists the required methods and algorithms they need to unveil new information from the data. In addition, SAP HANA’s text search engine allows the processing of unstructured information out of any kind of document and to convert it into structural information to be processed together with all the other data. The geospatial engine within SAP HANA allows representation and analysis of data along its geo-location parameters ensuring to track positions of products and assets which is e.g. very important for required service and maintenance scheduling.

It is not only important to extract all relevant information out of an SAP system, e.g. master data, historical service data etc. to make analysis on top possible. Even more important is that based on the findings in the prediction and analysis area, actions are directly taken and triggered in the back-end systems offering an end-to-end perspective. This transfers the predictive maintenance solution into a kind of “integrated online system”. Making it possible to schedule required service activities and technicians, or to look at the inventory and ensure required spare parts are available, or in case of production to integrate with quality management systems to define improvements in the design or development process.

Finally it is quite obvious that two parts of the solution require a very individual, and thus, customer-specific, flavor. The predictive algorithm needs to be developed based on information and experiences gathered from the customer probably over the last years. All this needs to be translated by the SAP data scientist into a mathematical, prediction model, to e.g. figure out new defect patterns. This of course requires dedicated alignment to understand customer’s business and requirements. The second area is the user interface. Keeping in mind the various industries and types of products or assets plus the different roles of those who work with the predictive maintenance solution, I don’t think there is a common user interface along the one size fits all approach. Hence building an individual UI based on SAP UI5 provides an user experience perfectly suited to their needs and working models. Our UI designers develop a concept, align it with the customer, provide several iterations of mock-ups and then, in a final user acceptance test, the UI is finalized tailored specifically to the end user needs and wishes.

To find out more about custom application development on SAP HANA, please visit

To report this post you need to login first.

1 Comment

You must be Logged on to comment or reply to a post.

  1. Waldemar Schneider

    I was right to focus on Predictive Maintenance for Use Cases on Big Data & Analytics based on project results at John Deere and Kaeser in IM&C and QatarGas in Oil & Gas.


Leave a Reply