The gap between knowing and understanding in the built environment
We have spent decades building sophisticated models of our buildings. We have BIM files, sensor networks, control systems, dashboards, and APIs connecting it all. And yet building owners are still asking a question that should have been solved years ago: why don’t my assets know what they are, where they are, or what they actually mean?
That question sits at the heart of a presentation by Kimon Onuma of ONUMA, Inc., delivered at the Asset Leadership Network webinar. His answer reframes the entire conversation around building technology, and it is worth understanding carefully.
The Meaning of Asset Intelligence in the Age of AI
The data is not the problem
Connecting data is not the same as understanding it. APIs don’t create meaning. Dashboards don’t explain relationships. Automatedbuildings This is a crucial distinction. The industry has spent enormous energy building pipelines that move data from one system to another, and the implicit assumption has been that more connectivity equals more insight. It doesn’t.
What has been missing is shared meaning. The ability for one part of a building’s ecosystem to understand what another part is actually doing, and why it matters.
The fragmentation problem hiding in plain sight
Building automation and control systems have always contained meaning. They know how systems operate, how relationships, sequences, and dependencies work. But that meaning has largely been encoded in system-specific logic, naming conventions, and structures that don’t translate easily beyond their boundaries. Automatedbuildings
Think about what that means in practice. Your HVAC control system knows everything about how air flows through the building. Your BIM model knows where every room is and what it is designed for. Your sensor network knows the real-time temperature of every zone. But none of these systems know what the others know. Each one understands its own world but not the building as a whole.
The result is a building full of intelligence that cannot talk to itself.
One building, many different realities
Part of what makes this hard is that every person who interacts with a building sees it differently. An occupant experiences a room as too hot or too cold. An architect sees it as a defined space with a function. A mechanical engineer sees it as a zone served by a particular air handling unit. A facilities operator sees it as a source of alarms. A landlord sees it as revenue per square foot.
They are all describing the same reality, but in different ways. Somewhere in between, meaning gets lost. Automatedbuildings
This is not just a technology problem. It is a communication problem that technology has not yet been designed to solve. Until now.
A real example: the cost of a poorly timed battery
In the PAE Living Building, the data was always there. Solar. Batteries. HVAC. Sensors. Controls. Everything was operating. But no one could see how it all worked together until it was connected. Once connected, simple issues became obvious, like a battery charging at the wrong time, creating unnecessary peak-demand costs. Not a design problem. Not a hardware problem. A meaning problem. Automatedbuildings
This is a striking example because it shows that the cost of fragmentation is not just theoretical. Buildings are making expensive, avoidable decisions every day because the systems inside them cannot see each other clearly enough to coordinate properly.
The Semantic Bridge
The proposed solution is something called a Semantic Bridge. It links the physical world of buildings, spaces, assets, and location with the semantic world of systems, relationships, rules, and dependencies. Select a room, and you see the systems behind it. Select a system, and you see what it actually serves. And when you ask a question, the answer isn’t guessed. It’s grounded. Automatedbuildings
This is a meaningful shift. Rather than building yet another platform that requires everyone to migrate their data into a new format, the Semantic Bridge connects what already exists. Bring your BIM. Bring your digital twins. Bring your building systems. If it connects, it belongs.
Why this matters for AI
AI doesn’t need more data. It needs structured meaning. Without meaning, AI guesses. With meaning, AI can reason. Automatedbuildings
This is perhaps the most important insight in the entire presentation. A lot of the current excitement around AI in buildings assumes that feeding more sensor data into a model will produce better outcomes. But raw data without context is not enough. If an AI agent cannot understand the relationship between a room, the air handling unit serving it, the sensors monitoring it, and the business rules governing it, all the data in the world will not help it answer the question “why is this room too hot?”
With structured meaning in place, that question becomes answerable. So do harder ones: Why did we hit peak demand at 2pm? Which systems are interacting in unexpected ways? What changed between last week and this week?
The role of owners
None of this happens unless owners require it. For decades, the industry has asked for deliverables: drawings, PDFs, disconnected models. What should be required instead is persistent IDs, connected relationships, and data that survives handover. The data already exists. It’s just not being required in a usable way. Until it is required, it will continue to be lost. Automatedbuildings
This is a call to action directed squarely at building owners, not vendors or technologists. The market will not fix this on its own. Owners need to change what they ask for at handover, and what they require from their service providers throughout a building’s life.
The bottom line
Buildings are not short on data. They are short on understanding. The difference between a building that reports a fault and a building that can explain why the fault happened, what it is connected to, and what the best response is, comes down to whether the systems inside it share meaning or simply share data.
That gap is finally starting to close. And once buildings can genuinely be understood, the work of improving them becomes possible in a way it has never quite been before.
