OPC UA Is Evolving Beyond a Protocol for Industrial Digitalization

The OPC Foundation, a Platinum Sponsor of ARC Advisory Group’s 24th Annual ARC Industry Forum in Bengaluru, was represented by Stefan Hoppe, President & Executive Director. In his presentation, OPC UA: The Foundation for Industrial Digitalization, Hoppe explored how secure semantic interoperability can support digital transformation across OT, IT, cloud environments, AI, digital twins, and connected value chains.

A central theme of the presentation was the need to look beyond the common perception of OPC UA as simply a protocol for transferring data between industrial systems. As organizations connect operational data with enterprise applications, cloud platforms, analytics, and AI, moving the data itself is only part of the challenge. Systems also need to understand what that data represents.

The OPC Foundation increasingly positions OPC UA as a secure semantic framework for industrial interoperability, combining standardized information models, flexible transport mechanisms, and built-in security.

Hoppe’s presentation can be viewed on YouTube or here:

<embedded link to be inserted>

OPC UA Is More Than a Protocol

OPC UA has traditionally been associated with communication between PLCs, SCADA systems, MES platforms, and other industrial systems. Hoppe argued that this view captures only part of its role.

Rather than defining a single communication protocol, OPC UA provides a collection of technologies for exchanging standardized information securely from the sensor and control layers through enterprise and cloud environments and applications such as digital twins and digital product passports.

The framework supports multiple transport mechanisms, including TCP/IP, UDP, and MQTT, as well as file transfer, REST interfaces, and the Open Web API. This allows organizations to select transport technologies according to the requirements of a particular application while retaining a common information model.

Security is also built into the architecture, covering areas such as information access, data transport, secure onboarding, and certificate management.

The distinction between transport and information becomes particularly important when technologies such as MQTT are involved. MQTT provides an efficient transport mechanism, but it does not by itself standardize the meaning and structure of the payload. “OPC UA over MQTT” can provide that semantic layer while using MQTT as the underlying transport.

From the Automation Pyramid to an Information Network

Traditional industrial architectures have often been organized as a hierarchy, with data moving through successive layers from sensors and controllers to supervisory, planning, and enterprise systems.

Digital transformation is making these boundaries increasingly fluid.

Industrial information may need to move horizontally between machines, vertically through OT and IT systems, or directly to edge and cloud applications. In some cases, information may bypass traditional architectural layers entirely.

The OPC Foundation describes this shift as moving from an automation pyramid to an information network. In this model, the priority is not simply establishing connections between individual systems but enabling standardized information to move securely across the industrial environment while retaining its meaning.

OPC Foundation illustrates the transition from a traditional automation pyramid to an information network supporting secure, standardized information exchange across industrial and enterprise environments

This model also creates a common foundation across OT and IT. The communication mechanism may change depending on where the information is being used, but the semantic description of the underlying asset, process, or measurement can remain consistent.

Keeping Context with Industrial Data

The value of industrial data depends heavily on context.

A temperature value, for example, becomes considerably more useful when a system also understands which asset produced it, the type of measurement involved, the engineering unit, its relationship to other equipment, and the operational process in which it is being used.

OPC UA information models are designed to provide this type of machine-readable context.

The OPC Foundation and its partners have developed more than 450 standardized information models, commonly known as Companion Specifications, covering industrial domains and assets including machines, robots, pumps, energy systems, buildings, field devices, and enterprise applications. These models are available in both human-readable and machine-readable formats.

The Foundation is also adapting these specifications for greater use with agentic AI. This work includes agent-friendly Markdown, textual descriptions of visual information, token-optimized retrieval-augmented generation (RAG) content, vector embeddings, and callable REST interfaces.

The objective is to make domain-specific industrial knowledge easier for both conventional applications and AI systems to consume.

Extending Industrial Semantics into the Cloud

Maintaining context becomes particularly important as industrial information moves into cloud environments.

A conventional approach may extract values from the shop floor and store them in a time-series database as timestamps, tag names, and values. While useful for many applications, this can remove relationships and contextual information already available in the OT environment.

The OPC Foundation Cloud Initiative is intended to preserve that context.

Its goals include improving interoperability between industrial and cloud applications; maintaining OPC UA information models in cloud environments; supporting AI analytics, industrial data spaces, digital product passports, digital twins, and industrial metaverse applications; and developing a common cloud reference architecture.

The initiative brings together major cloud and industrial technology suppliers. Participants and contributors highlighted in the presentation include AWS, Google Cloud, Huawei, Microsoft, and SAP, alongside automation suppliers such as ABB, Beckhoff, Honeywell, Mitsubishi, Rockwell Automation, Schneider Electric, Siemens, and Yokogawa.

The Foundation is also developing tools that make standardized information models easier to create and reuse. The OPC UA Cloud Library provides a repository of standardized industrial information models, while the NodeSetEditor allows manufacturers, machine builders, and system integrators to create structured models and move data in context between the factory floor, cloud environments, and third-party applications.

Digital Product Passports Put the Model into Practice

Digital Product Passports provide a practical example of why semantic interoperability can extend beyond conventional factory connectivity.

The OPC Foundation is working with CEN/CENELEC around Digital Product Passport standardization, with OPC UA providing a framework for describing information in a standardized, interoperable format.

Hoppe highlighted a Digital Battery Passport implementation in which OPC UA is used to model the information associated with a battery rather than simply functioning as a conventional client/server communication mechanism.

The information can originate during production and continue to be updated during the battery’s operating lifecycle. This creates the potential for a persistent digital record containing standardized information about the product and its history.

According to the presentation, implementing the Digital Battery Passport model required approximately half a day for the information model and another half day for the REST interface. The implementation is being made available as open-source software under the MIT license.

h

The Digital Battery Passport example demonstrates how OPC UA information models can connect production data, cloud services, and product lifecycle information within a common semantic architecture

The example illustrates a broader shift in the role of OPC UA. The technology does not necessarily require an OPC UA server and client in every use case. Its information-modeling capabilities can also provide standardized descriptions that other applications and interfaces can consume.

Preparing OPC UA for Industrial AI

Semantic information is also becoming increasingly relevant to industrial AI.

AI systems can work more effectively with industrial data when they receive not only values but also standardized descriptions of what those values represent and how different pieces of information relate to one another.

The OPC Foundation has established an OPC UA for AI Working Group focused on several areas, including supporting programming with OPC UA, analyzing OPC UA data, developing next-generation devices that use generative AI as an interactive interface, and applying AI to specification-development activities.

Another objective is to standardize interaction between OPC UA and generative AI using widely adopted technologies such as Model Context Protocol (MCP). The Foundation has also developed open-source MCP servers intended to make it easier for AI systems to interact with OPC UA environments.

Hoppe demonstrated the concept using a natural-language request to identify pH meters on a network and retrieve a sequence of measurements. Instead of requiring a user to manually identify devices, establish connections, retrieve individual values, and process the results, an AI interface can use standardized information about the devices and their data to perform the task.

The presentation also highlighted a tobacco-industry implementation involving Philip Morris International and suppliers. A common Tobacco Companion Specification provides standardized descriptions of equipment, giving end users a more consistent information foundation for integration and AI applications.

Building a Common Information Foundation

As industrial organizations connect more systems, cloud platforms, digital twins, and AI applications, interoperability increasingly depends on more than establishing communication between endpoints.

The receiving system must also understand the information being exchanged.

Hoppe’s presentation positioned OPC UA as a way to preserve that meaning across different technologies and architectural layers. Transport mechanisms can vary according to application requirements, while standardized semantic models provide a common description of industrial information.

This approach also helps explain why technologies such as MQTT, REST, cloud services, and OPC UA do not necessarily compete with one another. They can serve different roles within the same architecture.

As industrial digitalization expands from connected machines toward cloud ecosystems, digital product passports, digital twins, and AI-driven applications, standardized semantics can provide the common information layer needed to connect those environments without repeatedly rebuilding the meaning of the underlying data.