Building on the recent partnership announcement from SAP’s Franz Hero, our SAP and Microsoft teams have been working together to further develop a Joint Reference Architecture (JRA) that brings the best of both worlds together in the area of SAP Digital Supply Chain and Microsoft Azure.
The Joint Reference Architecture serves as a guide for customers and partners who are on their digital journey, showcasing optimal configurations and patterns to help navigate the best capabilities of both companies. This brings with it a choice in technology, optimized configuration of applications and services, and real-world extensibility scenarios.
As the portfolio of applications is spanning SAP Supply Chain Networks, SAP Digital Manufacturing Suite, SAP Intelligent Asset Management and SAP Integrated Business Planning, initial focus areas have been in the following areas:
Expanding cloud capabilities to the edge starting with SAP Digital Manufacturing Cloud and Azure Stack Edge
Now more important than ever on the factory floor
Leveraging Azure services to build additional value with applications such as SAP Integrated Business Planning
Such as a more accurate forecast that can drive manufacturing and fulfillment
IoT and Device Integration options with SAP Digital Supply Chain applications
Providing more choice and the ability to build on existing customer investments in IoT
The Joint Reference Architecture, represented above, is a result of the collaboration between our SAP Digital Supply Chain (DSC) team and Microsoft Azure teams to not only define a view for customers and partners on our approach, but as well to highlight engineering efforts taken by both organizations to define, develop, test and deliver innovations in edge and IoT. We reached a major milestone by the end of 2020, where we announced that our DSC portfolio of cloud applications, are now available in Azure data centers in Europe and North America. With the applications built on a technical foundation of both SAP Business Technology Platform and SAP IoT and Edge Services, and running on Azure means we can combine complimentary offerings to maximize value for our end customers.
Having multiple applications, technologies, and services can be difficult to represent in a single view since there are various data flows and connection paths between all of the many components. With this in mind, our goal was to provide an easy-to-read starting point that captures the main elements that we can further drill into with more specifics.
Now that we’ve covered the main highlights of the overall Joint Reference Architecture, let’s take a deeper look at a few of the scenarios.
Expanding Edge Capabilities
Edge computing is a hot topic among many organizations, especially within Digital Supply Chain, wherein the core concept is around having our workloads and processing closer to the sources of data. This is an important step in a manufacturing setting, since as the trend toward cloud computing continues, certain requirements (described later) are best met with edge solutions on the shop floor.
As part of the portfolio of applications within DSC, Digital Manufacturing Cloud (DMC) is a cloud-based manufacturing execution and performance solution linking production and business processes, with connectivity across multiple layers, from shop floor to enterprise systems. First released in 2018, DMC has now evolved to include DMC for edge computing, which offers a hybrid cloud-edge option for manufacturers. DMC is based on a micro-services architecture, and in order to realize the subset of the cloud workloads on the edge, our team partnered with the Microsoft Azure Stack Edge engineering team to simplify the landscape and operations for our customers.
The Azure Stack Edge appliance is offered as hardware-as-a-service and serves as the managed runtime environment for our DSC applications and related services. With a consistent and managed Kubernetes environment being made available, our development teams, along with customers, can avoid the complexity of setting up and managing such a cluster. Besides reduced setup time, we see the following benefits:
Security hardened solution
Low maintenance effort for customer IT
Minimal setup - only network, power supply, and location required
The following diagram illustrate a typical high-level deployment:
Edge nodes are centrally configured and operated by Digital Manufacturing Cloud (DMC) with each node running a set of services to handle the various aspects of data movement, e.g. Business Object Synchronization Service (BOSS), and life cycle activities, e.g. Edge Lifecycle Management (ELM) to name a few. These supporting services and operations are configured and managed from the cloud, so while the appliance is running close to the production floor, many of these tasks are handled with cloud qualities.
In such a setup, the core business application can now operate at the edge, with a subset of the cloud functionality available to plant operators and technicians. The end-users are accessing application screens that are running directly on the Azure Stack Edge device, with a synchronized copy of all relevant data objects necessary for manufacturing execution, such as the work orders, resources, routings etc.
Having Digital Manufacturing Cloud for Edge Computing running on such an appliance means that production facilities can extend the cloud processes close to production operations, and with this address the following challenges when dealing with cloud applications:
Need for real-time decision making
Data volume and bandwidth capacity issues
Decoupling of central and site operations
A solution that addresses these challenges, brings additional benefits in the form of local and efficient data processing, especially when dealing with high data volumes such as image processing for Visual Inspection. In many cases, these high-data volumes can be processed and generate actions or events without sending a bulk of the data stream to the cloud.
Lifecycle management remains flexible, and in control of the customer. Development also benefits, as a single code line can be maintained, meaning the services that are developed in the cloud can run on the edge without needing to be re-written or adapted.
Our teams have taken key learnings from our various beta deployments with customers and partners. Multiple network setups have been tested and are supported, and we continue with focus on delivering a scalable and robust offering that will grow beyond our manufacturing offering.
Building additional value with DSC Applications and Azure
Another key aspect of the Joint Reference Architecture revolves around combining Azure services with DSC Applications. With the DSC applications running in Azure Regions, we immediately see benefits when considering customer specific applications and services that may be integrated with SAP applications. By having applications and services within the same Azure subscriptions, customer can benefit from reduced latency and efficient data transfers. Identity and Access management is simplified and provides an improved user experience, building on the integration between SAP Cloud Identity Services and Microsoft Azure Active Directory.
A specific customer scenario is a great example of demonstrating this additional value, and we’ll take a deeper look at SAP Integrated Business Planning for Supply Chain (SAP IBP). Supply chain planning across the entire value chain network covers multiple domains, such as Demand Planning, Response & Supply Planning, Sales & Operations, Inventory Optimization and Supply Chain Control Tower for visibility. SAP IBP covers all of these planning domains across the strategic, operational and execution time horizons on a single platform complemented by intelligent predictive functions that are increasingly supported by AI and ML. It also provides for very open extensibility into other platforms such as Microsoft Azure that can provide customers a critical innovative edge on collecting and processing internal and external data that may influence the business plan.
In this customer case using the Azure services to collect and model various demand signals and feed these inputs into SAP IBP.
As depicted above, specific events are captured and processed using Azure services. Beyond the collection and storage of these events, the data is fed into a machine learning model developed and maintained using Azure Machine Learning Services, and the results of the model prediction form additional inputs into SAP IBP demand plan. With this additional data, the customer can respond to real-time customer demand signals (point-of-sale, orders, shipments) with pattern-recognition algorithms. This results in a more accurate forecast that can drive manufacturing and fulfillment, resulting in lower inventory and higher service levels.
IoT and Device Integration options with SAP Digital Supply Chain applications
Both SAP and Microsoft have strong offerings in the areas of IoT and device connectivity. With this in mind, we focused this area of our Joint Reference Architecture in providing choice to customers.
To add some details to the choice, customers have the option of (a) SAP IoT for Device Management and Connectivity and SAP managed Time Series Persistence or (b) Azure IoT Services for ingestion and Device Lifecycle management and Azure Data Explorer for Persistence.
The applications currently supporting this option are SAP Predictive Asset Insights and SAP Enterprise Product Development, with additional applications planned in the near term.
Our teams have produced some great reference material, including a sample application and templates located in our github sap-samples repository here.
More to Come
As our teams continue to develop and refine our Joint Reference Architecture, please watch this post for additional follow-up blogs where I’ll share be sharing more details around the current and upcoming scenario implementations.