The EU Data Act changes the conversation about machinery service
A machine may generate a continuous stream of operating data while its owner still struggles to obtain the information needed for a service decision. With the EU Data Act applying from 12 September, access to that information is becoming a more explicit part of the relationship between equipment users, manufacturers and maintenance providers.
As of September 26, 2025. The immediate opportunity is better-informed service. The harder task is turning accessible records into an accurate diagnosis, a compatible replacement and a machine returned safely to work. Those steps require different capabilities. Treating them as one automatic benefit risks disappointing equipment buyers and overstating what the new rules deliver.
What changed in September
The European Commission’s launch announcement describes new rights for users of connected products to access data generated through use and share it with third parties. Industrial machinery is among the relevant product categories. The intended effects include more choice in aftermarket services and opportunities to develop services around equipment data.
That makes the service discussion more concrete. A buyer can ask what information a machine generates, where it is held, how it can be obtained and whether an authorised maintenance partner can use it. For a supplier, a clear answer becomes part of explaining the product’s working life, alongside parts availability and service response.
The announcement establishes a policy change, not measured evidence that repairs have already become cheaper or faster. Those outcomes depend on the equipment, the available information and the organisations using it. A machine with accessible records can still wait for a specialist component or a qualified technician.
Access has boundaries and a separate design timetable
The regulation distinguishes access to readily available product and related-service data from a general entitlement to every piece of a manufacturer’s intellectual property. Its framework includes protections for trade secrets and personal data. Inferred or derived information produced through additional investment in analysis is not simply equivalent to raw or preprocessed product data.
Timing also matters. While the Act generally applies from 12 September 2025, the Article 3(1) design obligation applies to connected products and related services placed on the market after 12 September 2026. That is a separate milestone. It would be misleading to tell buyers that all existing machinery must already have been redesigned for direct access.
These boundaries are reasons to define the actual request. A maintenance provider seeking operating temperatures, fault events and timestamps should identify those records and the intended use. A request for proprietary diagnostic software, confidential design files and unrestricted control access is a materially different proposition. Contract and product-specific assessment remain necessary.
A data file needs an explanation
An exported value is only useful if the recipient understands it. An illustrative temperature record may identify a sensor but omit its position, unit, sampling interval or the operating state of the machine. A service engineer cannot confidently compare it with another machine’s reading without resolving those details.
This is where data access becomes an engineering task. Buyers should ask for definitions, units, time conventions, equipment identifiers and relevant configuration information. They should also understand whether a value is measured directly, filtered, estimated or aggregated. Those distinctions affect what can reasonably be inferred from a trend.
The OECD’s February 2025 policy brief on data access and sharing provides a broader explanation. It describes a continuum of access arrangements and emphasises complementary resources, including skills, software and interoperable metadata. It does not establish a return on investment for industrial maintenance. Its relevance is that access alone is only one ingredient in useful reuse.
Repair choice still depends on physical capability
A user may obtain a fuller operating history yet remain dependent on a limited supply of bearings, drives or specialist assemblies. Data cannot manufacture a spare part, establish its compatibility or provide the skill needed to fit it. The procurement discussion should therefore keep information access and physical service support connected but separately specified.
Consider an illustrative production machine with recurring stops. Better records might help a service company distinguish a mechanical issue from an operating-condition problem. If replacement is required, however, the team still needs the correct revision of the component, suitable installation procedures and a way to verify the result. A data-sharing agreement does not answer those questions by itself.
This suggests a useful comparison when evaluating suppliers. Ask each to describe a realistic fault-to-recovery sequence: who receives the records, who diagnoses the issue, who confirms the part, who performs the work and who accepts the machine back into service. An attractive promise about open data becomes more meaningful when those responsibilities are visible.
Industry collaboration is growing, but announcements are not outcomes
Siemens’ 23 September announcement of a data alliance with machine builders, research and industrial partners shows another direction for machinery data. The stated purpose is to exchange engineering, manufacturing and machine information for industrial generative AI applications. It is an announced collaboration, not proof that every participating machine now offers a finished service or that customers have achieved a particular saving.
Its relevance to equipment owners is the expanding number of possible users and uses for industrial information. Data collected for one service relationship may also be valuable in product development or model training. Those purposes should be described clearly rather than bundled into an ambiguous permission to use everything.
An owner considering a voluntary arrangement can ask what is shared, with whom, for how long and for which purpose. It can also ask how withdrawal, retention and commercially sensitive production information are handled. These are practical governance questions, distinct from the specific rights and obligations established by the Data Act.
Start with one service problem
A sensible first project is a bounded maintenance problem on a defined equipment group. Identify the decision that better data is expected to improve. It might concern when to inspect a recurring fault, whether operating conditions explain an alarm or whether two service providers reach the same diagnosis from the same evidence.
Then test the entire information route. Can the user retrieve the records when needed? Can the receiving technician interpret them without repeated clarification? Are gaps and clock differences visible? Can the service finding be linked back to the particular machine and configuration? This exercise can reveal practical barriers before the business commits to a broad platform programme.
The test should have an outcome measure that fits its purpose. Reduced time spent obtaining a usable record is different from reduced downtime. A pilot may improve the former without yet demonstrating the latter. Keeping those measures separate prevents a promising technical result from turning into an unsupported commercial claim.
Make service terms part of the equipment decision
The OECD’s earlier research on data sharing argues that there is no single optimal level of openness for every context. That is relevant to industrial service, where useful access must coexist with security, confidentiality and responsibility for physical equipment. More data shared with more parties is not automatically a better arrangement.
A machinery purchase should therefore include a practical discussion of data availability, retention, access continuity and support. Buyers should understand what happens if a service contract ends, a machine changes owner or a service provider stops operating. These are commercial continuity questions that can be addressed before an urgent breakdown makes negotiation harder.
The September change creates an opportunity to make machinery service more transparent and contestable. Real value will come from usable information, competent diagnosis, available parts and clear responsibility for restoring production. Equipment suppliers that can explain that whole chain will offer something stronger than a generic claim that their machines are connected.