Four tips for designing scalable IIoT networks

Don’t get distracted by the promises of the latest so-called solutions for Industrial Internet of Things (IIoT). Focus on these essential design principles while building a data network. Intelligent input/output modules and other tools can add flexibility.

By Josh Eastburn July 5, 2021

 

Learning Objectives

  • Open technologies can be used in IIoT projects.
  • Integrate cybersecurity in the IIoT design.
  • Process data at the edge, use distributed I/O modules.

So your boss wants to take on an Industrial Internet of Things (IIoT) project, but you’re not sure what that means? Or maybe a pilot project is partway through and running into obstacles? In either case, don’t worry. Focus on one word: Scalability. Intelligent input/output (I/O) modules can help with scalable, flexible and secure IIoT designs.

Scalability is the property of a system to grow and manage increased demand, and it is what distinguishes IIoT from previous architectures for data and control. Even if working on a small-scale application, focusing on scalability will clarify critical technology decisions, removing obstacles from the path and creating a foundation to build on.

As you build out an IIoT network, follow these four tips to maximize scalability.

ONE: Take advantage of open technologies for IIoT

During the same time that the telecommunications industry created the modern internet, the controls industry went through the fieldbus wars, taking years to accept open standards like Ethernet on the plant floor. Proprietary communications media and protocols have created complex networks of point-to-point connections between data consumers and data producers that are expensive to scale up.

While this isn’t the only reason for the difference in the scale of consumer and enterprise information technology (IT) systems when compared to industrial networks, adopting efficient, vendor-neutral standards has been a benefit to scalability in those domains.

Openness can benefit IIoT scalability in a number of ways:

  • Open standards promote interoperability, which is fundamental to the concept of IIoT.
  • Open system architectures reduce vendor compatibility as a design constraint, allowing engineers to optimize for performance, instead.
  • Open technologies often reduce licensing costs, providing more funding for system expansion.

Message queuing telemetry transport (MQTT) is an example of an open IoT standard that is scalable and interoperable, especially when paired with Sparkplug B payloads. Many resources are available to learn about MQTT. Sparkplug B helps with data integration among sensors, devices, gateways and applications in an MQTT infrastructure.

TWO: Integrate cybersecurity in the IIoT design

A network cannot expand if it can’t be done safely. The lack of integrated security in traditional automation is another factor that has limited its scalability. It tends to either push IT groups toward draconian security policies that inhibit basic IIoT goals, like collecting data from separate automation networks, or increase the complexity of the network further through the use of additional cybersecurity devices and software.

Like communications, cybersecurity benefits from the use of standard technologies and techniques, rather than proprietary solutions added to automation products as an afterthought. User authentication, SSL/TLS encryption, and VPN, to name a few, have been common in IT devices for a long time and have started appearing in newer OT devices, as well. This also is a boon to IT interoperability and trust-building.

Security is fundamental to IIoT and should be given high priority when selecting network components. While not a security protocol, MQTT owes its popularity in part because it leverages proven mechanisms like TLS encryption and allows IIoT devices to communicate bidirectionally while remaining closed to outside connections.

THREE: Process IIoT data at the edge, distributed I/O modules

After designing a high-performing network, it can be tempting to focus the system on transporting data to cloud services for processing and analysis. This is a common objective for early IIoT projects. However, raw data streams are expensive to transport, store, and process, and may result in the untimely demise of a project due to budget overruns.

A more scalable approach uses edge-oriented processing to streamline the quantity and improve the quality of data sent to central processing services. Processing power can be embedded at the site or area levels but also can be distributed down to the line, cell or device level thanks to recent developments in industrial edge I/O and control.

Edge computing power is typically used to filter noise from raw signal data, aggregate data from multiple sensors, add contextual metadata, or package time-series data into interoperable exchange formats like JSON. All of this effort at the edge provides clean data throughout the system and reduces later communication and number-crunching costs.

FOUR: Manage the IIoT device lifecycle

Anything limiting the ability of the system to grow in response to demand should be considered in the system’s design, and the physical layer of an IIoT network is full of potential bottlenecks. Networking constraints have already been touched on, and due consideration should be given to the lifecycle of IIoT devices as well.

For a truly scalable system, the following, at least, will need to be quickly repeated and maintained over time:

  • Device selection or specification
  • Physical installation, including appropriate environmental protections
  • Connections to power, communications, and I/O or data sources
  • Device configuration
  • Configuration of device, data, and network security
  • Integration with other systems.

Small aside: When discussing IIoT, engineers sometimes assume it requires wireless connectivity and smart sensors, but this is not true. The majority of valuable industrial data is found in traditional wired devices and sensors. Understanding how to incorporate this brownfield data into the network can unlock significant scale.

Keep IIoT simple: Edge services at the I/O level

The key to cutting through the hype surrounding IIoT is understanding simplicity is a prerequisite for scale. Plant-floor communication is nothing new, but the traditional technologies for plant communication create too much complexity to achieve the order of magnitude increases in connectivity that are the goal of implementing IIoT.

Fortunately, an increasing number of options are available for building scalable IIoT networks at a manageable cost of ownership. By moving connectivity, security and data processing capabilities further down the automation stack, they reduce the industrial networks’ complexity and the effort required to scale them up.

Edge-oriented services embedded down to the I/O level can be used to integrate legacy devices, bridge disparate automation networks, create segmented security zones within the IIoT network, and provide efficient processing and transport directly to shared systems.

Many so-called IIoT solutions promise turnkey results at a high price without improving on the fundamental constraints of the previous generation of technology. Don’t get distracted. Following these tips will help in applying essential principles of system design and help get that IIoT project done.

Josh Eastburn is director of technical marketing, Opto 22. Edited by Mark T. Hoske, content manager, Control Engineering, CFE Media, mhoske@cfemedia.com.

KEYWORDS: IIoT project design, intelligent I/O modules

CONSIDER THIS

Have you considered the latest technologies as part of scalable IIoT design?


Author Bio: Josh Eastburn is director of technical marketing, Opto 22. After 12 years as an automation engineer working in the semiconductor, petrochemical, food and beverage, and life sciences industries, Eastburn works with the engineers at Opto 22 to understand the needs of tomorrow's customers. He is a contributing writer at blog.opto22.com.