The installation industry consists of work that is hard to digitize because it is physical: a technician who has to be on site, a fault that only becomes visible once the cabinet is open, a part that is not available while the customer waits. The hours go into work preparation, planning and dispatching, into taking in and distributing service requests, into reporting and post-calculation, and into being on the road itself. What drives the outcome is not just the complexity of the technology, but the extent to which information about installations, stock, and history has been recorded somewhere in a way a system can read. Where that recording is missing, the work stays human work, even when the task itself is simple.
In some installation companies, planning is now supported by systems that automatically assign service requests based on location, urgency, and available technician, with a planner who approves or adjusts the assignment. That is the second category: AI makes the proposal, a human assesses it with reason. In other companies, that assignment is still done entirely manually, not because the technology is missing, but because requests come in via phone, email, and app side by side, without that information coming together in one fixed place. The difference, then, is not in ambition, but in whether the underlying data has already been made usable.
Remote fault diagnosis shows a similar picture. Some companies have a system read along with sensor data or earlier fault reports first, to suggest a probable cause, after which the technician takes that assessment into account or dismisses it. That shortens the time on site without replacing the technician: the diagnosis remains checked by someone who actually sees the device. Where that sensor data does not exist or is not accessible, the fault still starts from zero, even if the model would already work elsewhere.
Reporting and post-calculation are the third area where the shift is visible. Work orders, hour registration, and material usage are, in some companies, automatically merged into a draft invoice that an administrator checks before it goes out. That is capacity being freed up, not capacity disappearing: the check remains, only the tallying is no longer manual. In companies where work orders still arrive on paper or sit in separate systems, that step does not exist, simply because there is nothing to merge.
Actually installing something, assessing a situation that does not fit the protocol, and the conversation with a customer about a solution that is not standard: that remains human work in every company, now and in the scenario most management teams envision. Not because AI could never support that, but because the liability, the physical action, and the context on site should rest with a human. An ambition that assumes this kind of work runs remotely or automatically requires a different kind of readiness than an ambition limited to planning and administration: different data, different authorities, and a different way in which technicians and planners work together with each other and with a system.
Management teams in this industry regularly formulate an ambition such as "less travel time" or "faster response to faults", without pinpointing what needs to be in place first: unambiguous asset registration, a fixed place for service requests, or authority for a planner to follow an automatic proposal without consultation. That step is often missing not out of unwillingness, but because no one has made it explicit. The ambition assessment captures that layer: from vision to a concrete target state, tested against readiness, with a translation back to who is allowed to make which decision if a system's proposal does not match what the technician sees.
The question of which work in this specific company can genuinely be taken over by AI, with which part under supervision and which part not, is answered per task using the work scan from FTE TO AI, independent of what already works elsewhere in the industry. Similar shifts are playing out in healthcare, where scheduling pressure and registration demand a different kind of readiness, in wholesale, where inventory and order processing form the bottleneck, and in the transport sector, where planning and routing face bottlenecks comparable to those in installation work.
This shift affects not only planning and administration, but also how a management team prepares decisions and communicates them to the shop floor. Anyone wanting to know how a change lands in practice before it is fully rolled out will find pointers in the early signals that show whether a change is truly taking hold, and anyone wanting to involve technicians and planners without ending up in a long-running process can read how that can be done more compactly at involving the shop floor in a strategy without setting up an endless process. Decisions about the personnel consequences of such a shift fall outside the scope of this discussion; separate legal requirements apply to those.
The free readiness check consists of eight short questions, one per dimension, and produces a picture of where your organization is furthest along and where it is not yet. The full ambition assessment, with the four layers and the five confidence gates, is under construction.
Vertel wat u wilt bereiken, dan kijken we samen wat daarvoor moet staan.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.