Editor’s note: FM Perspectives are industry op-eds. The views expressed are the authors’ and do not necessarily reflect those of Facilities Management Advisor.
If you manage a big portfolio of buildings, you know exactly who I’m talking about. These are the veteran engineers who can walk into a mechanical room, listen for half a minute, and immediately point out which chiller is short-cycling. They remember that the rooftop units on the east wing of Building 7 always go out of whack when it’s humid. They know which valve sticks after a long weekend and that it’s faster to just fix it by hand than wait for a work order. None of this is in a manual. It’s all in their heads, built up over decades of actually doing the work.
The industry calls them building whisperers, and they are disappearing.
The experienced building engineers who carry this kind of institutional knowledge are overwhelmingly in their 60s. The pipeline of younger replacements behind them is thin to nonexistent. This is not a new observation. Anyone who manages commercial buildings at scale has been watching this trend build for years. But what has changed is that the tools to do something about it are finally catching up to the problem.
The Knowledge Gap Nobody Budgets For
When one of these experienced engineers retires, you’re not just losing a person—you’re losing decades of pattern recognition. The new hire might be sharp, but they’re starting from zero when it comes to the building’s quirks, history, and all the ways things can go wrong. Now scale that up to hundreds or thousands of buildings, and it’s easy to see why a lot of operators say this workforce transition is their No. 1 operational risk.
Most places try to solve this with documentation. Write up the procedures, keep the service logs current, build out the training manual. I get it. But I have talked to enough building operators to know that what is in the manual and what actually keeps a building running are two different things. One of the stories that was told to us was about a veteran engineer who could walk past a mechanical room and know something was off just from the sound. His replacement had the same certifications, same tools, same documentation. But he did not have 20 years of knowledge about that specific building and its peculiarities. That is the kind of knowledge that does not transfer through a PDF.
AI as an Institutional Memory
This is where I think edge AI is starting to show up in a way most people didn’t expect. It’s not focused on replacing the skilled engineers. It’s about finally capturing and surfacing the kind of operational know-how that used to only exist in someone’s head.
By processing live data from building systems right on-site, edge AI can watch for patterns, spot anomalies, and connect the dots across systems in real time. So now, when a less skilled tech walks into a building, they can ask the system for context that used to require a 30-year veteran standing next to them. What’s this unit’s normal range? When did it last act up like this? What fixed it last time?
This is not speculative. Large building operators are already using AI to auto-tag and normalize operational technology data in minutes, work that used to take integrators weeks and cost tens of thousands of dollars per building. When you remove that bottleneck, you also remove one of the biggest barriers to getting new staff productive quickly. They no longer need to manually learn every system’s naming conventions, quirks, and data formats before they can start reaching well-informed decisions.
Building AI for the Workforce You Actually Have
There is an important nuance here that the industry is still working through. Most of the conversation around AI in buildings focuses on models, algorithms, and analytics platforms. But the real limiting factor in many buildings is not the sophistication of the AI. It is the skill level of the people who need to use it.
A former colleague of mine at a major technology company used to say, “If you solve the ordinary, dumb problems, people will pay you all day long.” The same principle applies here. Facility teams do not need AI that requires Python scripting or command-line prompts. They need a button that says “show me all the set points” and a dashboard that builds itself from the query. They need interfaces customized to the skill levels of the people actually on-site, not the skill levels of the engineers who built the system.
That’s why AI tools need to have solid safeguards built in. Systems have to prevent accidental data loss with confirmation steps, use role-specific access controls, and provide workflows that help less experienced staff get to the right solution, even if they don’t have deep technical backgrounds.
The companies that figure this out, that build AI for the workforce they actually have rather than the workforce they wish they had, are the ones that will win in building operations over the next decade.
Stretching Expertise, Not Replacing It
The real benefit here is not so much about reducing headcount, but rather about enabling the same team to manage a greater number of buildings with fewer difficulties. Providing facilities teams with AI-powered edge tools offers improved visibility and reduces the requirement for manual workarounds. The experienced engineers who remain become even more valuable, as their expertise helps to inform and improve the AI. Meanwhile, new staff are no longer starting entirely from scratch; they are entering a system that already contains much of the building’s operational history.
This is not the story that makes headlines. The headline version of AI in buildings is always about energy savings, autonomous operations, or substituting human decision-making entirely. But the quieter, more practical reality is that AI is becoming a bridge across a workforce gap that has been widening for years. It is preserving hard-won knowledge, rendering it available to people who did not earn it the hard way, and helping building operators keep running as their most experienced people walk out the door.
The building whisperers earned their expertise over decades. The question is no longer whether AI can replace them. It is whether we are smart enough to capture what they know before they are gone.
Chris Timmins is chief commercial officer at IOTech Systems, where he leads the company’s building automation strategy.