Why can’t you trust your asset data, and who owns the fix?

Your PLM holds the latest engineering status. Your ERP holds financial and procurement data. Your CMMS holds maintenance history. Sensors deliver condition data through IoT. Your document management system (DMS) and your digital and physical archives hold drawings and documents. Each is right from its own perspective, but nobody has the full picture.

In practice, people choose the source they can reach fastest. That is rarely the most current or the most reliable one.

The quality of a decision is limited by the quality of the data it is based on, however good the dashboard. Consider a replacement decision based on an outdated condition score. Replace too early and you spend capital on an asset that still performs well. Replace too late and you risk an unplanned failure and everything that comes with it.

The same applies to maintenance planning, investment prioritization and compliance reporting. When people do not trust the numbers, they verify them by hand, which makes every decision slower and more expensive. AI is no exception: a model fed with conflicting data produces predictions that no one trusts, as we explore in Why do so many AI initiatives in asset management fail to deliver?

Replacing every system with one platform is expensive, slow and rarely necessary. The aim is to connect the data from your existing systems into one trusted record per asset: a record that shows what is true at any moment, whatever the source.

Two things make this practical. The first is the asset register: the structured list of what you own, where it sits and how it is classified, which every other system refers to. The second is a common naming structure: a reference designation system (RDS), for example based on ISO/IEC 81346, that gives each asset one designation across your PLM, ERP, CMMS and sensor data.

That requires clear rules. Which system is the master for which data? How is information linked across systems? What happens when two sources disagree? Without answers, every improvement in maintenance, compliance or analytics stays built on sand.

Many organizations launch a data cleanup or digitization project, and many see the same problems return. Data reflects what the organization needed at the time it was created. Without processes, requirements and ownership, it decays again at the next change, project or system migration.

Fragmentation often starts in investment projects, where new assets and documentation enter the environment without clear requirements and a controlled handover. Securing new data at the point of creation stops the problem from growing while you improve what you already have.

Data governance means deciding who owns your data and what rules keep it correct. Too often it is left to IT, as a matter of systems and access rights. But the questions are about the business. Who is responsible for the accuracy of this data? Who can approve a change? What quality standard applies to the data behind an investment decision?

That is why ownership of data belongs with leadership. They are the ones accountable for both the business value of the data and the systems that hold it. They can set one clear standard for who owns what and what quality is required, so each project and department does not have to invent its own.

The ISO 55000 series points in the same direction. ISO 55001 expects top management to make sure the organization determines what asset information it needs, keeps it under control and defines roles and responsibilities. The logic is simple: decide what information you need and who owns it before you choose tools.

An AIM Assessment covers the first step. You get a clear view of your current information landscape, a gap analysis, and a roadmap with a recommended first step. The roadmap also shows you how to move on to steps 2 and 3, so you know what to do next and in which order.