Manufacturing is not an office environment with machines added on. It is a sector in which a physical process — raw material, processing, product — is the leading factor, and in which digital systems always follow behind it or have to ensure that the process keeps running. That order determines what AI maturity means here. An ERP system, an MES, sensors on a machine: they exist to support the physical process, not to function independently. That is fundamentally different from professional services, where the work itself is already digital and data is a starting point rather than a derivative.
That layering makes measuring maturity more complicated. A manufacturing company can have a modern planning application and still have no idea what the machine on the shop floor is actually doing, because that data has never been made accessible. Conversely, a company with outdated software can still have good insight into its process, because people on the shop floor have been recording accurately for years. Maturity and the modernity of systems do not automatically go hand in hand here.
Within the seven dimensions used by the maturity measurement, a specific pattern often stands out in manufacturing. IT infrastructure and data management — two of the fundamental dimensions — regularly show a large gap between production and office. On the shop floor, systems run that are built for reliability and uptime, not for sharing data. In the office, systems run that are indeed built for exchange, but are not always connected to what happens in the factory. Those two worlds often use different definitions of the same term, and that happens without anyone identifying it as a problem until an AI project gets stuck on it.
The organizational dimension has its own pattern. Decisions about investments in production processes often lie with people with a technical background, while decisions about data strategy lie with others. When those two groups rarely sit at the table together, an organization emerges that looks ready for AI on paper, but in practice follows two separate agendas. That is a different spread than in education, where the gap more often lies between policy and the daily practice of the professional, not between two operational pillars within the same company.
The plot round — in which several people score the seven dimensions independently of each other — often exposes a sharp contrast in manufacturing between production management and the IT or executive layer. A production manager sometimes rates the data management dimension highly, because the machine records exactly what happens. An IT manager rates that same dimension low, because that data is nowhere structured or made accessible for analysis. Both observations are correct; they are simply looking at a different part of the process.
That spread is functional, not uncomfortable. It shows where a shared picture is lacking within the organization, and that is exactly the information needed before anything is added to a pilot. A pilot built on the fundamental dimensions — organization, infrastructure, data — without that spread having been noticed, risks getting stuck on a problem that no one had identified because no one had the complete picture.
Because the physical process is the leading factor, the temptation in manufacturing is often to point AI directly at an operational problem: predicting maintenance, checking quality, optimizing planning. Those are visible, attractive applications on the dependent dimensions. But whether they work depends on what needs to be in place beforehand: is the data from the machine recorded in a way that is reusable, is there someone responsible for the quality of that data, and is there an organizational structure in which production and IT share the same definitions. Without that foundation, a pilot remains an isolated experiment, however good the model may be.
That order does not only apply here. In the transport sector, a similar tension plays out between operational systems and central data, and in the agricultural sector the physical process — the crop, the animal, the season — is equally the leading factor over what digital systems can record and influence. Manufacturing is no exception to this logic, though it is a sector in which the consequences of a skipped foundation quickly become visible on the shop floor.
The maturity measurement does not pass judgment on a factory or a sector; it produces a picture of five levels across seven dimensions, filled in by several people from the organization itself. That picture shows where the fundamental dimensions already stand and where they are still lacking, and where opinions within the company diverge. That is a different conversation than whether a pilot is impressive; it is the conversation about whether the organization can carry a pilot.
Once it is clear where the organization stands on the dimensions that need to carry AI, a different question arises: which part of the work itself is eligible for transfer to AI. That is not what this measurement answers. That is what the work scan from FTE TO AI is for, which calculates per task what part of it can be taken over by AI. The maturity measurement and the work scan connect to each other: one shows whether the foundation is in place, the other shows what changes in that work, given that foundation. Anyone who combines both pictures not only knows whether the organization is ready, but also where the first change will concretely present itself.
The maturity measurement from hybridresourcing is under construction. Anyone who wants to complete the plot round as soon as it becomes available can sign up for the waiting list.
Vraag maar wat er moet staan voordat AI in uw organisatie kan landen.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.