The Human Layer: Understanding The Automation Paradox In Property And FM

• Content
• Smart Buildings & Technology
• The Human Layer: Understanding The Automation Paradox In Property And FM

16 April 2026 | Updated 15 April 2026

The hype around AI in FM should not be allowed to obscure the essential human element, says Joss Rennocks, Transformation Director at Bellrock.

AI is changing facilities management right now, delivering smarter buildings, fewer breakdowns, tighter energy performance and a better user experience, all with less manual intervention.

But there’s a paradox hiding inside this progress: the more we automate, the less humans touch the system day-to-day.

That means engineers stop ‘doing’ and start ‘monitoring’. Callouts reduce. Faults become less visible. Dashboards become the interface. It feels like maturity - until something goes wrong.

When that happens, it becomes painfully obvious that automation hasn’t removed the need for expertise, it has just concentrated it. It has moved competence away from routine tasks and into fewer, higher-consequence moments where judgement matters most. This is what I call an ‘automation paradox'. And in the FM sector, it’s not theoretical.

I believe that AI will make FM better, but only where human expertise is protected, developed and empowered. The winners in the next era won’t be the firms with the most sensors or the brightest dashboard. They’ll be the ones with the strongest human layer around the technology.

FM isn’t a clean environment for automation

Facilities management isn’t a factory line. It’s not a controlled process with predictable inputs. FM is messy, with ageing assets, layered upgrades, mixed occupancy patterns, systems that no longer match drawings, and user expectations that shift faster than most plant can respond to.

This matters because automation assumes something that FM rarely gives you: consistency. Data points go missing. Sensors drift. Control logic gets modified over time. Overrides are applied and forgotten. Asset information is incomplete or simply wrong. And yet, the building still needs to function, safely, compliantly, and comfortably every single day.

So when people say “AI will replace engineers,” they’re looking at FM through the wrong lens. Engineering in FM isn’t only about reacting to faults. It’s about interpreting complexity, managing risk, and balancing trade-offs that are rarely obvious from a dashboard.

The Automation Paradox: why expertise becomes more important

Automation is fantastic at reducing routine work. It removes repetitive actions, automates decisions, stabilises control, and even predicts failure modes before the human eye sees them. But in doing so, it erodes competence.

Engineers build their competence through repetition. It’s the difference between somebody who can reset a plant and somebody who instinctively knows the chain of causality underneath it. If automation removes the repeated exposure to system behaviour, competence can quietly erode. Over time, the engineer becomes a supervisor of systems rather than an owner of systems. Which is fine - until an incident hits a variable that the algorithm hasn’t seen, or the data doesn’t represent reality. And FM is full of variables.

Automation fails when variables collide: when the building is being used differently; loads shift; something external changes the conditions; or the control strategy is technically correct but practically wrong for the human beings occupying the space.

AI can detect issues, but detection isn’t the hard part anymore. The hard part is interpretation, action and accountability - and these are all human processes. 

Let me ground this in a real example. We worked on a multi-tenant 300,000 sq ft office building. The landlord managed the main services (boilers, chillers etc.), but tenants owned the terminal units that provide heating and cooling across the floor plates - a common set-up that’s also a perfect recipe for ‘not my problem’ attitudes if it’s not managed properly. The issue wasn’t dramatic, just persistent, painful instability: uncontrolled and inconsistent heating, causing constant, justified complaints.

In this type of scenario, it’s easy to fall into the comfort trap: keep tweaking setpoints, chasing individual complaints, altering schedules. It looks like action, but it never stabilises the system. It’s reactive theatre.

The turning point came when we used an analytics platform and saw something important: phantom demand for heating. The building was asking for energy without a valid reason. Digging deeper, we found the culprit: failed-open FCU heating control valves that were allowing constant LTHW flow through terminal coils.

This was the automation paradox in the wild. The platform highlighted the symptom, but the resolution required human engineering thinking, and a willingness to tackle the system, not the complaint.

We changed the BMS strategy. We repaired the valves. We flushed and rebalanced sections of the circuit. We didn’t treat comfort like a dial, but like an agreement, training the tenant team to shift culture around what is genuinely acceptable in an operational multi-tenant building. 

The impact was immediate and measurable. But AI didn’t deliver those outcomes by itself - people did. The technology provided visibility. The expertise created stability. And the relationship management made it stick. That’s the human layer.

In the age of automation: maintenance becomes more important, not less. Smart diagnostics don’t compensate for neglected physical plant.

So what is the human layer in practical terms? It’s not more people. It’s not nostalgia for reactive FM. And it’s definitely not anti-technology. The human layer is the capability that turns AI from insight into outcomes. It’s the difference between a building that looks good on dashboards and one that performs reliably in real life.

It needs competence - not just qualifications, but real technical depth. It also needs relevant training, engineer empowerment, and meritocracy, where performance is measured in outcomes, not activity.

So as AI shapes the next phase of facilities management, as it undoubtedly will, don’t assume that this means FM needs less human expertise. It means FM needs better expertise.

The role of the engineer is evolving. It’s no longer just fixing plant. It’s interpreting imperfect data, judging risk, and intervening decisively when systems behave outside the model. In the age of AI, the real differentiator isn’t who has the most data. It’s who has the capability to turn that data into reliable outcomes, everywhere, every day.

Picture: A headshot of Joss Rennocks, Transformation Director at Bellrock.

Article written by Joss Rennocks | Published 16 April 2026

Share

---

---

Related Articles

---