Data exchange standards in production: why companies are still not ready

Data exchange standards in production bring order to communication between machines, MES, ERP, SCADA, and analytics applications. They do not prove that a company can share data safely or use it in a way that supports decisions. After reading this article, you will be able to check whether your organization is ready for open data exchange, or whether it only has a few integrations that work because specific people know how to keep them running.

This topic matters most to production directors, maintenance leaders, automation engineers, IT/OT managers, and people responsible for plant digitalization. If you already use MES, ERP, SCADA, or plan communication through OPC, you should know whether production data gives you decisions you can trust.

OPC Foundation describes OPC as an interoperability standard for secure and reliable data exchange in industrial automation. OPC UA builds on that direction as a platform-independent architecture. OPC DA still works in many older environments, especially where infrastructure relies on Windows and COM/DCOM.

That gives companies a good technical base. Machines, controllers, supervisory systems, and applications can exchange data without writing communication from scratch every time.

Communication alone does not help much when data has no context. A machine signal must connect with a batch, work order, recipe, shift, operator, quality status, and event history. Without that, you see measurements, not the process.

A simple readiness test starts with one question: can you trace data from a sensor to a business decision? If answering it requires calls to several people, the integration works only at a technical level.

Production data now affects cost, quality, and delivery performance. It feeds OEE reports, downtime analysis, batch tracking, production planning, maintenance, and energy use control.

Eurostat reports that in 2025, 46.45% of EU enterprises used ERP systems. In large companies, the share was 88.71%. BI systems were much less common, used by 16.28% of enterprises. The numbers show a gap: companies often have tools for managing processes, but less often have an analytics layer that helps them work with data in a consistent way.

In manufacturing, that gap quickly affects daily work. ERP stores product, batch, and order information. MES describes production flow. SCADA and PLCs provide line signals. Spreadsheets often become temporary registers for corrections, exceptions, and comments.

When these sources are poorly connected, the company starts reconciling reality by hand. A downtime report differs from a shift report. Quality data cannot be linked to process parameter history. The MES and ERP integration works, but teams still argue about definitions.

Data exchange standards in production reduce dependence on one vendor and bring order to communication between OT and IT. Each standard covers a different part of the need.

AAS has a specific role where a company wants to describe an asset in a standardized way. IDTA publishes Asset Administration Shell specifications covering the metamodel, API, data based on IEC 61360, and security. This matters when data exchange covers more than signal reading and must include component descriptions, documentation, parameters, and relationships with other systems.

A good data architecture does not start with choosing a standard. It starts with the decisions that data must support. Only then does it make sense to discuss OPC UA, OPC DA, AAS, data sharing through OPC, or MES-ERP integration.

The largest gaps usually concern data meaning. A variable named Temp_01 may be read correctly through OPC, but it tells little to anyone outside automation.

Ask what the value actually describes:

Time is another weak point. Production data becomes useful when it can be tied to a specific event. A delay of a few minutes between systems can distort downtime, failure, or quality deviation analysis.

Security needs the same attention. An internet-wide study of OPC UA deployments found many configuration issues: missing access control, disabled security functions, outdated cryptographic mechanisms, and shared certificates. The standard can support secure communication, but a poorly configured environment still exposes the company to risk.

Responsibility is often unclear. Automation defines tags, IT maintains servers, production uses reports, quality interprets deviations, and controlling calculates costs. When no one owns the shared data model, every machine or system change starts another round of manual checks.

The EU Data Act has applied since September 12, 2025. It concerns access to data generated by connected products and related services. For manufacturing, this is a clear signal: data from machines, devices, and systems is no longer only a technical record of equipment work.

Production companies will face more questions about what data is created, who can use it, what can be shared, and how access is protected. This concerns machine manufacturers, machine users, integrators, and service providers.

Without a data map, answering these questions can take time and money. With a data map, it is easier to assess what can be shared, what must stay restricted, and what needs extra technical or legal conditions.

Choose one process that has a strong effect on cost, quality, or delivery. Good candidates include downtime, batch tracking, OEE, process parameter control, or sending work orders from ERP to MES.

Then ask the team these questions:

If the team cannot answer all of them, the next integration will most likely repeat current gaps. Clear answers mean you can choose the standard and scope of work with more control.

The best first step is a data flow map. It should show sources, systems, owners, read frequency, dependencies, and the decisions supported by data. Such a document quickly shows where the company loses consistency.

Only after that check should the team design the integration. Otherwise, it may build a technically correct connection that still fails to meet the needs of production, quality, or planning.

At explitia, we can help you map data flows, set up communication through OPC, integrate MES and ERP, and organize architecture for open data exchange. The article can include subtle internal links to materials about AAS, OPC UA, OPC DA, data sharing through OPC, and MES and ERP.

Data exchange standards in production are necessary, but they do not replace order in data. A company ready for open data exchange can name sources, owners, definitions, risks, and decisions that depend on production information.

The best test is not the number of connections between systems. Check whether shop-floor data helps people make decisions without manual explanations of where it came from and what it means.

If the answer is uncertain, start with one process and a data map. This small step often protects the company from an expensive integration based on wrong assumptions.

They are agreed ways of communicating, describing, and sharing data between machines, control systems, MES, ERP, SCADA, and analytics applications. Common examples include OPC UA, OPC DA, and AAS.

New projects more often use OPC UA because it is platform independent and supports a broader information model. OPC DA still runs in older environments, especially where infrastructure is based on Windows and COM/DCOM.

No. Open data exchange means organized and controlled access. The company still manages permissions, security, data scope, and responsibility.

Start with one process, such as work orders, batches, downtime, or OEE reporting. The team should define data sources, field definitions, owners, and the decisions that the data must support.

Check it when reports need manual corrections, systems show different values for the same process, or knowledge about integrations depends on single people. These are signs that the company needs to organize data before the next implementation.