What does interoperability mean in industry?

Interoperability describes two or more systems exchanging data where that data has the same meaning on both sides. The second part of that sentence is the hard part. Establishing a connection is rarely the problem today; making sure a transmitted value means the same thing in the target system as in the source almost always is.

In production, this problem meets a particularly heterogeneous landscape: controllers of various generations, control systems, MES, ERP, sensors, cloud services, and analysis models — built up over decades, with different goals and different remaining service lives.

Interoperability is therefore the silent prerequisite of almost every initiative filed under Industry 4.0 or Smart Factory. Connectivity, digital twin, predictive maintenance, cross-plant metrics — all of them assume that systems understand each other. Anyone who starts these initiatives without clarifying this layer pays for it again in every single project afterward.

This distinction comes from information architecture and reliably explains what a project actually hinges on.

• Technical

  The systems are physically and protocol-wise reachable. Network, transport, ports, encryption. Projects fail here least often.

• Syntactic

  The data has a shared format: data types, units, structure, timestamps. A value in degrees Celsius and a value in Kelvin are technically the same and syntactically not.

• Semantic

  The data has a shared meaning: which object, which state, which reference quantity. This is where it is decided whether an evaluation across two plants is valid.

• Organizational

  It is clarified who operates the interface, who approves changes, and who is responsible in case of a fault. Without this level, none of the other three holds.

Six systems, two architectures: fifteen maintained connections — or six integrations.

No single standard solves all four levels. The choice follows the task, not the trend.

Companion Specifications: where transport becomes meaning

OPC UA alone defines how data is transmitted in structured form, not what it means. This gap is closed by Companion Specifications — extensions developed by industry associations that define, for a given machine type, which information is available under which name and in which unit.

The practical effect: two injection molding machines from different manufacturers deliver the same metric under the same name. Without this agreement, the same value is named differently by every manufacturer, and the job of standardizing falls to the user — once per manufacturer, per plant, per project.

For procurement, this translates into a concrete question for every machine supplier: which Companion Specification does the controller support, and in which version? It belongs in the tender, not in commissioning.

The German reference framework: RAMI 4.0 and the Asset Administration Shell

In German-speaking countries, RAMI 4.0 — the Reference Architecture Model Industry 4.0 — organizes these building blocks: it spans hierarchy levels, lifecycle, and architecture layers as three axes and shows where a given standard applies. The Asset Administration Shell
(IEC 63278-1) is the carrier of the characteristics within it, and ECLASS is the dictionary for it.

You don’t need this model to connect a machine. You need it to justify, in a tender, why a particular interface is being required.

Semantics and ownership remain a separate determination in any case. A Unified Namespace is the place where they become visible.

• Effort grows quadratically. Six directly coupled systems produce fifteen connections, ten produce forty-five. Via a shared layer, it is six or ten, respectively.
• Data silos. Values exist, but only within one system. Analyses get built in spreadsheets and go stale immediately.
• Dependence on individuals. Self-built couplings rarely survive staff turnover.
• Vendor lock-in. Anyone who integrates only via proprietary routes no longer freely decides on the next plant — or freely on price.
• Delayed initiatives. Every new analysis starts with an integration project instead of a business question. What that means for IT response time

Until recently, interoperability was a voluntary architecture decision. Pressure is now coming from two directions, and both concern the question of who is allowed to access machine data.

• Data Act. Regulation (EU) 2023/2854 governs access to data generated by the use of connected products. For manufacturing companies, this mainly means: data from your own machine is no longer automatically kept by the manufacturer. Anyone who wants to use this data needs a structure for it — otherwise the entitlement stays theoretical.
• Data spaces. Under terms like Manufacturing-X, Catena-X for the automotive supply chain, and Gaia-X as a framework, cross-industry spaces are emerging where companies share data under controlled conditions. Participation requires exactly what is usually missing internally: a named, semantically unambiguous structure of one’s own data.

Both developments shift the task: interoperability used to be the prerequisite for your own analyses; it is becoming the prerequisite for taking part in your own industry’s data exchange at all.

It gets presented as a technical topic, and it is an architecture decision with long-lasting effect:

• Longevity. Plants run 15 to 25 years, IT systems 3 to 5. The connectivity layer has to outlast both.
• Scalability. Whether an initiative makes it from the pilot plant to the whole group depends on how reusable the integration is.
• Time to market. How quickly a business idea goes into production depends less on the idea than on the number of interfaces that have to be opened for it.
• Technological independence. Open standards keep the next procurement decision open.

pronubes is the layer between the shop floor and IT: it captures production data securely, translates it into a shared form, and brings decisions back into the processes as instructions.

• pronubes Edge runs locally in the plant and translates between OPC UA, MQTT, REST, SQL, and file formats — and buffers when the connection fails.
• pronubes Zones assigns the structure in which a value unambiguously belongs to a site, area, and unit — the semantic layer, operationally maintained.
• pronubes Insights makes connections, data streams, and system states visible; an interface whose failure nobody notices is not one.

Over 100 system types can be connected — from current IoT protocols to decades-old interfaces. More about the platform