Skip to Content
Business Trends

Four Keys to Success with Digital Supply Chain and IoT

Part 1 of 2 in the Digital Supply Chain Architecture Success series.

If you’re a manufacturer in the B2B space, you’ll have noticed a trend by now: increasingly, customers evaluate you on the basis of the experience you deliver not just the products you produce. Are you easy to work with? Do you help your customers achieve their goals? Do you require customers to adjust to you – or the other way around?

To deliver experiences that lead to repeat business and ongoing loyalty, manufacturers are redesigning their value chains. The goal is to serve customers better. The way forward is to better understand customer needs and build in capabilities for responding to those needs with agility.

This requires data – which is where the Digital Supply Chain and Internet of Things (IoT) technology come into play. With IoT sensors that report on the health of deployed assets, devices, or machines at customer facilities, manufacturers can pull in customer data that helps them not only respond faster but predict what’s coming next.

The architecture of success

If you’re on the IT or operations team for a manufacturer, the question remains: how do you actually implement an IoT architecture capable of driving Digital Supply Chain success? At SAP we’ve put a lot effort into designing a common reference architecture focused on four key capability areas. To succeed with IoT and the Digital Supply Chain, you need to:

  1. Connect things
  2. Process and analyze data
  3. Persist and manage data
  4. Optimize processes

Let’s dive in.

Connect things

IoT for the Digital Supply Chain starts with connecting devices deployed in the field with home base. The goal is to ingest device data and move it along for processing and analysis. Device sensors can either connect directly to home base via the cloud or they can communicate with an edge gateway that lives in close physical proximity to the deployed devices forming a local network.

This edge gateway can process defined amounts of data locally and make decisions about when to send to home base. This can be useful when real-time data is not required or when a live connection may not be available – as might happen with, say, a deep-sea oil rig exposed to inclement weather.

Whether transmitted through an edge gateway or directly via the cloud, data ultimately lands at home base – which, in our case, is the SAP Cloud Platform. Raw sensor data is brought in through a secure and highly scalable cloud gateway that acts like a front door.

For device data to get in through the front door, interoperability is paramount. For example, while MQTT is perhaps the most widely known protocol for transmitting IoT data, specialized protocols are increasingly common from industry to industry. Your cloud gateway needs to support them all. It also needs the openness to integrate  with device management solutions provided by third-party partners.

For data received, the IoT platform then provides critical business services – such as persistence, streaming, machine learning, blockchain, and business transaction processing. In cases where low latency and near real-time reactions are required, some of these services can also be delivered at the edge, where relevant data can be filtered and pre-processed as needed before being sent to home base.

Process and analyze data

Incoming data needs to be ingested, processed, and semantically enriched – all duties managed by SAP Data Hub. This hub applies rules on data streams, raises events, and performs the tasks of data transformation and cleansing. It also enriches raw operational data from devices with business data from ERP, supply chain management, asset management, and other critical systems.

Our architecture also calls for a flow-based data pipeline paradigm. SAP Data Hub helps out with tools to create and modify data pipelines using flexible operators. It also supports a highly scalable, server-less approach with open-source container orchestration using Kubernetes. Data federation, meanwhile, helps you connect to customer data lakes without full replication – making it easier to bring in data from outside.

The point of processing is to prepare data for meaningful analysis. More than ever, this analysis can be performed by non-data scientists using intuitive tools. SAP Analytics Cloud, for example, gives lines of business capabilities for reporting, dashboarding, data exploration, and simulation.

Users also have access to the SAP Leonardo Machine Learning Foundation, which provides tools for creating algorithms – as well as prepackaged options that let you get moving right away. From analyzing device data for predictive maintenance to optimizing logistics with live-streamed vehicle sensor data, business people can now put their data to work without the delays typically associated with engaging data experts.

Persist and manage data

IoT implementations require a smart approach to persisting and managing big data – one that balances cost and efficiency with access and availability. Our architecture calls for a tiered approach based on the notion of cold, warm, and hot data stores.

Cold store is reserved for most raw data. This data is stored on servers in the cloud (using Hadoop Distributed File System or Amazon S3, for example). It is cost efficient but slow.

Some raw data, however, needs to be more accessible – such as recent sensor output. This is where warm store comes into play. In our architecture, warm store is supported by SAP Vora – which combines resilience and fast access with high flexibility and moderate costs.

Hot store, finally, is for your most important and frequently used data. Based on SAP HANA in our architecture, hot-store data lives in memory and typically contains aggregated and pre-processed time series data enriched with higher business semantics. Here is where operational technology (OT) and information technology (IT) converge – a requirement when it comes to linking devices to business processes.

Optimize processes

After combining machine and business data for analysis and insight, now it’s time to turn this capability back on your business processes. The goal is to drive process improvements – and even new ways of doing business. To these ends, our architecture prioritizes the creation of targeted IoT solutions based on reusable microservices that are open for partner and customer development.

SAP Cloud Platform provides a strong platform-as-a-service offering for building these microservices in a multi-cloud context. With a rich set of technical platform services made possible with SAP Cloud Platform and a wide range of SAP Leonardo Intelligent Technologies, you have a leg up on developing IoT applications that drive process improvements.

In addition, many IoT solutions are directly embedded into core SAP applications such as SAP Manufacturing, Warehouse or Transport Management. This makes it easier to integrate your IoT deployments into your business. And with support for digital twin technology, you can create an asset intelligence network that maps your entire IoT deployment (and others as well) to the digital world. This enables you to analyze data from a greater pool, speed analysis, and introduce product improvements faster at lower risk.

Stay tuned

All of this adds up to an architecture with data management and solution creation capabilities designed to help you take fast advantage of everything IoT technology has to offer. But the question remains: how are organizations using this architecture to realize results in the real world? For stories on exactly this, stay tuned for our next installment in this two-part series.

Be the first to leave a comment
You must be Logged on to comment or reply to a post.