In the world of Industry 4.0, integration is key. If you're looking to make your production sites more efficient and productive, it can be challenging to connect every component in your infrastructure. Luckily, reference architectures can help you with a framework for those connections. In this article, we'll explain what IIoT reference architectures are, why they matter, which components they consist of, and how those components communicate with each other.
‍

What are IIoT reference architectures?

Industrial Internet of Things (IIoT) reference architectures are used as a blueprint for designing digital systems within an industrial or manufacturing environment. They can assist organisations in their digital transformation journeys by helping them connect the OT and IT sides of their production sites, while avoiding the typical pitfalls of these integrations.
‍

Why should you use an IIoT reference architecture?

IIoT reference architectures are helpful in providing insight into how different parts of their infrastructure interact with each other, allowing organisations to identify areas where improvements can be made. This connecting factor is especially important when we consider the difference in longevity between OT and IT components.  

OT hardware tends to be used for many years in largely the same configuration to ensure continuity and consistency in production. IT, on the other hand, is characterised by frequent software updates and hardware replacements to keep up to date with the latest innovations. By connecting both ends with a central broker, IIoT brings innovation to OT without sacrificing on reliability.  

The connections also make it possible for less tech-savvy end users to create dashboards through a Manufacturing Execution System (MES) or low-code applications like Microsoft Power BI. This improves readability compared to SCADA visualisations, which are more technical in nature and require in-depth knowledge of the underlying process.
‍

What does an IIoT architecture look like?

An IIoT reference architecture consists of an IT and OT node layer connected to an intermediary broker, which sends its data to a SCADA or IIoT host. The connections between the IT nodes and the broker are referred to as northbound connectivity, while the broker’s connections to OT nodes are referred to as southbound connectivity. Both connectivity lanes are used bilaterally.

The IT layer contains application nodes that represent connections to different parts of the IT infrastructure. The possibilities include the MES mentioned above, a data historian, various analytics tools (including AI and machine learning), and a wide range of other applications. Since these systems tend to be relatively new, they can directly interface with the broker.

On the OT side, there are various devices, sensors, and Programmable Logic Controllers (PLCs) that can send data to and from the broker through the same protocol. In reality, not all of these components will be able to connect directly because of differences in age, technology, and protocols. That’s why, unless the entire OT setup is (relatively) new, this side of the architecture will typically include additional gateways that support the new messaging protocol.

‍

How can the components of an IIoT architecture communicate with each other?

There are different connectivity options, and some organisations even prefer to develop their own standards. The reference architecture that we use is based on two tried-and-true conventions: a Unified Namespace (UNS) and Sparkplug.

Unifying the namespace in an IIoT reference architecture means using the same naming conventions throughout your organisation. You will no longer have to wonder which of your many sensors any given one is. A UNS also allows for easier and more efficient communication within and between systems, because it lets components use the same language across different networks and devices. This means that any data can be shared quickly and accurately between all of its components.  

Sparkplug also helps enable data interoperability between different edge devices in an IIoT architecture. Sparkplug is an extension of the MQTT protocol that is specifically geared towards industrial data. It allows for the bidirectional flow of real-time data between nodes that we mentioned earlier, and offers some advanced features for manufacturing environments: report by exception, store & forward, low bandwidth, high scalability, low latency, ...

Using Sparkplug, edge devices such as sensors and PLCs will be able to easily communicate with each other (and even with cloud systems), without having to think about transport protocols. We chose Sparkplug because it has seen widespread adoption and development throughout the industry thanks to its flexibility, although it is still less well known in Europe at the moment.

Besides communication, using the same naming convention means that the collected data is collected and stored consistently, which can help companies make more informed decisions about their IIoT investments. However, this does risk compromising the data security. That’s why we have a dedicated partnership with a security firm to ensure that any vulnerabilities are detected and removed before moving to production. Sparkplug also ensures end-to-end reliability and further enhances security by providing authentication and encryption for all nodes within the IIoT system.