Interoperability at the Edge: How to Build a Modular Process Control System

In 2022, a global survey asked more than 2,000 IoT developers about the biggest barrier to IoT’s growth. Interoperability was the third most common response, with 22 percent of respondents naming it the major concern. This is where the need for building modular process control systems (PCS) comes in.

Edge computing was also becoming a lot more common in industrial IoT. There were around 638 million edge-enabled devices in enterprise IoT in 2022—a figure that’s expected to grow to over 1.2 billion by 2030.

Taken together, these trends suggest that we’ll struggle to realize the promise of IIoT until we establish universal standards for IoT edge systems. To better understand this dynamic, let’s take a closer look at a central element of industrial automation systems: PCS.

An Incompatible Industrial IoT Edge

Process control technology has evolved over decades to address the needs of countless end applications. Inevitably, the different requirements of varied process control use cases resulted in an ecosystem of incompatible solutions produced by technology suppliers operating in silos. 

Eventually, this made it difficult to upgrade or migrate a process control deployment without a complete teardown and rebuild. Just ask the control systems engineers at ExxonMobil. The oil, gas, and chemical provider has been at the forefront of process automation technology for decades. If there’s a barrier on the road to Industry 4.0, the ExxonMobil team has faced it.

Around 2018, the team’s big challenges boiled down to a lack of flexibility. Without at least limited interoperability, you couldn’t mix and match components.    

“There was a lot of frustration at the time,” said Steve Bitar, then-R&D program manager at ExxonMobil. “We were dealing in a world of industrial control systems where all the devices were proprietary.” 

How do companies like ExxonMobil work toward standardizing IoT edge computing, networking, and storage components so process control systems can be incrementally upgraded? How might you create your own flexible, updateable IIoT system? More to the point, what’s required to usher in true plug-and-play capability for industrial IoT at the edge?

Here’s what it will take to build a modular PCS system that allows you to select from a variety of interoperable solutions that best suit your needs.     

Why the Process Industry Needs Modular PCS Architectures

In 2018, it was hard to improve one thing about a PCS without replacing the whole system. 

“You had to migrate the entire control system to change anything,” Bitar said, recalling the challenge of 2018. “You’d have loved to upgrade one element or another, to get more processing power or more memory. Maybe you had a cybersecurity vulnerability you wanted to close, but it would require swapping out all the hardware.” 

Many PCS users are left grafting new technologies onto old systems, leading to unwieldy, ad hoc solutions. 

“You’d just end up layering defense on top, like adding firewalls and work processes to protect your equipment,” Bitar said. “You ended up boxed in.” 

One manufacturer’s compute module wouldn’t work with another’s I/O device. One power supply wouldn’t plug into another switch module. You couldn’t optimize your control system beyond what your vendor had to offer.  

These conditions didn’t meet the needs of industrial operators, and they still don’t. 

“What we want is plug and play,” said David DeBari, control systems engineer at ExxonMobil. “You plug in your mouse, you plug in your keyboard, and they work well. I want the same thing with my controllers.” 

Two related design principles can deliver on that plug-and-play promise: interoperability and interchangeability. 

The Two Key Characteristics of Modular Process Control Systems

“There are two sides to plug and play,” DeBari said. “You have to have a connector everyone agrees upon, and you have to have protocols—a canonical model of behavior and the parameters we now call data models.” 

Data models and connectors are concrete illustrations of interoperability and interchangeability. Here’s a closer look at these important concepts for modular process control systems:   

• Interoperability allows multiple components in a computational system to exchange meaningful data with one another. The key word here is meaningful—in an interoperable system, all components understand one another. 

This requires a common method of data representation. It requires a common format for communication. In other words, it takes common specifications. Interoperability occurs at the software and firmware levels.

• Interchangeability, on the other hand, refers to hardware. It’s a question of connectors, the busses, and ports and plugs. Devices that share the same connection technology are interchangeable.

To create a truly modular PCS, every component must be interoperable and interchangeable with the rest of the system. You need both.

The only way to achieve interoperability and interchangeability is for the market to agree on common specifications.

“Interoperability is going to require a standard, agreed-to method for being interoperable,” DeBari said. “We need to agree on data representation and a communication method. You also need interchangeability: Does this connector fit? If we don’t agree, it doesn’t work.”  

Ideally, the specifications that standardize PCS components will be open to all industry stakeholders. Open IoT specifications create more choices, more innovation, and more favorable economic conditions for buyers. They generate market forces and channel them toward solving all the challenges of Industry 4.0.

Open standards for IoT edge devices and PCS architectures are already available. They proved a big part of the solution to ExxonMobil’s challenge.

Open, Modular Standards for Process Control Systems

The InterEdge specifications from PICMG—a nonprofit consortium that develops open standards for industrial IoT—define a fully modular PCS architecture. InterEdge proved to be part of the solution for ExxonMobil’s flexibility challenge, along with open standards from the Open Process Automation Forum (OPAF), another standardization group.  

Today, any IoT manufacturer, systems integrator, or control systems engineer can access InterEdge. This open standard provides specifications for:

• Compute modules
• Switch modules
• I/O modules
• Power supply modules
• Bus communication protocols
• Data models built for process control

Interedge specifications define physical characteristics for modules, including electrical requirements and pinouts. That keeps the system interchangeable. Meanwhile, InterEdge’s data models ensure interoperability. 

Interedge features multiple layers of abstraction, starting with the behavioral models at the firmware level. Crucially for physical process automation, these behavioral models cover both sensors and actuators; that is, they support data harvesting and control for automated machinery at once. 

Collaborative Data Models

Above the firmware level, InterEdge uses the same communication protocols and data modeling techniques as the Internet (without exposing IoT systems to the World Wide Web, of course). These protocols are ubiquitous and well-known, so they simplify development for industrial IoT systems. 

PICMG’s open specifications for industrial IoT avoid reinventing the wheel in another way, too. The consortium partners with the Distributed Management Task Force (DMTF)—another standardization organization—to build functionality into the DMTF Redfish application programmer interfaces and schemas.

This partnership between PICMG and DMTF led to Redfish data models like automation nodes and control points, which support fully synchronized robotic motion control with motion profiling. Another PICMG Redfish data model enhances verification for jobs assigned to individual machines, ensuring that, say, a CNC machine is fitted with the right bit before starting the automated job sequence. 

The significance of these events is simple: Multiple standardization organizations — OPAF, PICMG, DMTF, and more — are coming together to construct interoperable, interchangeable systems for industrial IoT. 

“Our philosophy at PICMG is that we develop things that need to be developed,” said Douglas Sandy, CTO at PICMG. “And we leverage what already works. We collaborate with other organizations where we can, and it’s been a really good collaboration with DMTF for Redfish.”   

It’s no accident that modular IIoT begins with collaboration. Ultimately, the goal is for IoT manufacturers to build devices that work with one another right out of the box. Consensus is both the goal and the means of reaching that goal. 

Once we reach a consensus on edge IoT standards, multi-vendor interoperability and interchangeability will simply become the norm. That’s the market ExxonMobil, PICMG, DMTF, OPAF, and many others are working to build—and it’s exactly what you need to build a modular, multi-vendor PCS—and other open and interoperable edge computing platforms that will move the IoT market forward.