At this year’s SAP Tech Ed, Internet of Things was front and center. Customers presented use cases during the keynotes, SAP personnel led numerous IoT technical sessions and labs offered developers the chance to connect smart devices to Hana Cloud Platform. For me, Tech Ed provided the perfect venue to dig deep into SAP’s extensive technology and business plans for IoT. 

SAP does not see IoT as a stand-alone market, technology or business venture.  Instead they see IoT as an incremental data source for their portfolio of business automation solutions. For customers, IoT augments the information and decision-making resources that SAP already provides. So SAP sees IoT as less “brave-new-world” and more “better, faster, cheaper”. For me this isn’t surprising; SAP has been a major IT provider to manufacturers worldwide and they have years of experience managing manufacturing data, including shop floor machine data.

SAP has devised an impressive stack of technology from which to build an IoT solution (See figure 1).  Very little of that technology is IoT specific, but is instead multipurpose, enabling SAP to leverage existing technology investments for IoT.

Figure 1.

A suite of IoT applications intended to improve physical asset operational ROI sits on top of the SAP’s IoT stack. These include:

• SAP Predictive Maintenance and Service - This solution monitors asset conditions, predicts machine health, proactively orders spare parts, improves spare parts planning, manages technician staffing.
• SAP Networked Logistics – This solution monitor traffic towards and within a hub in real time, Enhance traffic volume and efficiency, decrease wait times for customers, reduce environmental impact.
• SAP Connected Manufacturing – This solution supports Industry 4.0, provides more holistic production status visibility, integrates production data with business process data, enables more responsive manufacturing.
• Other connected device applications – These solutions connect and enhance a variety of operations including cities, oil-fields, construction and utilities.

SAP will use these applications to maximize their value to customer. That value will come from improved management and use of assets and facilities instrumented with smart sensors. And that value will serve multiple functions throughout their customers’ organizations from R&D to manufacturing to service and support.

To optimize the deployment and use experience of those applications, SAP offers the Hana Cloud Platform (HCP) with all of its services, tools and APIs. HCP is the beating heart of their IoT solutions, just as it is for their other next-generation business applications. Amongst all of the services and abilities provided by HCP, my particular interest is in its analytic capabilities. My view is that the technology closest to decision making is the most influential, which in many, many cases is analytics. The good news and the bad news is that analytics are to be found throughout HCP. It’s good because customers have lots of options in their solution design; it’s bad because optimizing a configuration will require some serious consideration of the trade-offs between the various options. Analytic abilities throughout SAP’s IoT stack provide data transformation/aggregation/filtering, metadata definition and use, data discovery and data visualization. SAP’s analytics-capable components include:

• Cloud for Analytics - Cloud for Analytics includes BI capabilities of reporting, dashboards, data discovery and visualization. Also includes budgeting/planning, and predictive model development/deployment. Cloud for Analytics will be formally launched mid-November 2015.
• Event Stream Processor - Event Stream Processor uses a streaming-oriented SQL-derived query language that’s capable of data definition, aggregation, joining and filtering. Obviously it has time-series abilities.
• SQL Anywhere/SQL lite - These RDBMS products provide smart device-embedded data definition, management, aggregation and other transformation capabilities.
• HANA, SAP ASE, SAP IQ - The columnar-style HANA and IQ RDMBS products as well as SAP ASE general purpose RDBMS provide data management, access and transformation capabilities.
• Business Objects - As Business Objects (BO) can access data from SAP’s various data management systems, so can it provide analytics for IoT. BO is not part of HCP at the moment, so runs on-premise.
• Manufacturing Integration &  Intelligence (MII) - MII provides metadata driven dynamic reports, dashboard visualization, SPC/SQC/6Sigma and other statistical/arithmetic analytics specifically for manufacturing. MII runs on NetWeaver and is an on-premise application.

A unique need of IoT analytics solutions is to connect to devices. That obvious requirement has generated no end of industry standards, implementations, etc. To be fair, one-size-does-not-fit-all when it comes to device connectivity. SAP rightfully offers multiple ways to connect and interact with smart devices:

• Message Management Services (MMS) – MMS is a new IoT-specific way for HCP to connect to smart devices. MMS supports RESTful connections using HTTPS and Web Sockets. In the future, MQTT will be supported as well. Data is transferred from devices to HCP record-by-record via JSON packages and uses OAuth tokens to authenticate data transmission. MMS will also persist data from smart devices on HCP.
• MobiLink -  MobiLink is a data synchronization service between SAP’s embeddable SQL Anywhere RDBMS and other databases including those available on HCP. Smart Devices would collect sensor data in SQL Anywhere tables that would then be replicated to HCP in batches.
• Plant Connectivity (PCo) – PCo is a framework and set of services and management tools that enable communication between SCADA-style systems, controllers, files, data historians and SAP manufacturing applications. PCo communicates on top of TCP and uses publication/subscription style message transfer.
• 3rd Party IoT connectivity – OSISoft provides the PI Integrator to export sensor data from its PI System to SAP HANA. Other 3rd party IoT data connectors will be provided in the future.

As part of SAPs’ IoT option for HCP, remote device management (RDM) will be offered to onboard/sunset, configure, command and extract data from smart devices. Both RDM and MMS are managed via the IoT Services Cockpit on HCP.

Future SAP-based IoT solutions will gain incremental functionality as HCP expands to include various open source technologies including those from Cloud Foundry. This is just one proof point of the value that SAP places on open source and passion-based developer ecosystems. Any developer can now apply for an account on Hana Cloud Platform and, once approved, start developing HCP apps with help from documentation, tutorials, and code samples on GitHub and community support.

Because of SAP’s long experience with manufacturing information technology automation, IoT is just an extension of what they’ve already been doing. But SAP does not treat IoT casually. They are very aware of the potential for IoT to dramatically reframe how business value is generated by manufacturing and other physical asset-centric industries. Their technology portfolio for IoT is a rich combination of past, present and future development. Choosing which SAP technology components for an IoT Solution to use will require deep knowledge of tradeoffs for each alternative. This will be a task best done with help from SAP-knowledgeable experts.