What the PAE Living Building Reveals About the Governance Gap in Commercial Buildings
Part 1 of 3: The Missing Binding Layer — The Story and the Problem
The Substrate That Was Always Needed
In February 2026, the Coalition for Smarter Buildings (C4SB) demonstrated the Interoperable Building Box (IBB) System at the PAE Living Building in Portland, Oregon. Presented at AHR/ASHRAE in Las Vegas, the demonstration illustrated—with striking clarity—the structural gap that the buildings industry has lived with for decades: capable systems that work individually but lack the governed relationships they require to function as a whole.
That gap is precisely what CNS/CP—the Connectivity Naming System using Connection Profiles—is designed to fill. Developed by Project Arete and defined as the inherent connectivity substrate of the IBB System architecture, CNS/CP provides the binding layer that enforces relationships between building systems structurally, not procedurally. The PAE building is now being prepared for the Arete implementation that will make this architecture operational.
The PAE Living Building: Where the Opportunity Became Clear
The PAE Living Building is one of the most advanced commercial buildings in the United States. Designed by PAE Engineers, it incorporates sophisticated HVAC, lighting, energy management, and occupancy systems—all functioning as designed. Yet after occupancy, a familiar pattern emerged—not because of any shortcoming in the building’s design, but because the industry simply has no infrastructure to preserve the relationships between systems after handover.
As Kimon Onuma documented in AutomatedBuildings.com, the building’s systems worked individually but could not see each other as a whole. During design and construction, thousands of engineering decisions shaped how energy flows and systems interact. After handover, that logic—as it does in every building—flattened into files. The Revit model became PDFs. Schedules lived in spreadsheets. Dashboards did not connect to each other. This is not a PAE problem; it is a universal industry condition. No building, regardless of how well designed, can avoid it because the binding substrate does not yet exist.
What distinguished PAE was recognizing this gap and choosing to address it proactively. The issue was not performance. It was the absence of governed relationships between systems
—a gap that PAE’s engineers recognized precisely because the building was advanced enough to make the missing substrate the only remaining barrier. The rationale for system interactions—the why behind each connection, the roles each system played, the context in which data exchange was meaningful—was scattered across tools with no structured binding. In a less capable building, this gap would have gone unnoticed. At PAE, it became the catalyst for action.
An Industry-Wide Pattern
This pattern is not unique to the PAE building. It is the defining structural challenge of the commercial buildings industry. Every high-performance building faces the same gap: systems can communicate via protocols like BACnet, Modbus, and increasingly through cloud APIs, but there is no standardized mechanism to establish why two systems should connect, under what terms, in what roles, and within what operational context. The industry has many capable endpoints and many communication protocols—but no binding substrate that governs the relationships between them.
What the C4SB Team Accomplished at PAE
The work completed at PAE over the past year represents the foundational preparation that a system governed through CNS/CP requires. The C4SB team—spanning NIST, SkyCentrics, CloudBIM (Kimon Onuma), and multiple working groups—accomplished a series of steps that, viewed through a governance lens, are precisely the prerequisites for governed connectivity:
• The Revit BIM model was translated into a machine-readable knowledge graph using the open-source BIM2RDF tool, aligned with ASHRAE 223P—establishing the semantic foundation for system relationships.
• CloudBIM provided an operational context layer with persistent identifiers, ensuring that equipment, spaces, and system relationships could be resolved across tools and over time.
• SkyCentrics introduced a “Discover” function that automatically identifies endpoints and associates them with the shared semantic structure—reducing endpoint mapping from months to minutes.
• A live demonstration at AHR 2026 showed real data flowing through the integrated stack.
This work established the landscape that CNS/CP is designed to support. The semantic alignment, persistent context, and automated discovery are not alternatives to CNS/CP—they are the preconditions for it. With the Arete implementation now being prepared, the PAE building’s owner will be able to use CNS/CP as the substrate through which they govern connectivity across their building’s systems—not governed by the technology, but empowered by it.
Three Governance Gaps No Existing Tool Fills
The PAE experience illustrates three structural challenges that every high-performance building faces. These are not technology gaps—they are governance gaps that no existing protocol or platform addresses.
The Discovery Problem. In today’s buildings, discovering what systems and capabilities exist across a facility is a manual task—tracing through commissioning documents, BIM models, and vendor dashboards. When a new analytics platform or energy management service needs to connect, identifying the right endpoint and verifying its capability is an engineering project in itself.
The Negotiation Problem. Declaring a capability is not the same as offering a guarantee. Today, a BAS system may advertise that it provides temperature data—but there is no standardized way to verify what that means, in what format, under what conditions, or with what reliability.
The Context Problem. As building operations span multiple systems, vendors, and management domains, context—operational policies, safety boundaries, authorization scope—tends to drift or disappear entirely.
Why Existing Approaches Don’t Fill This Gap
Practitioners will reasonably ask: don’t we already have tools for this? BACnet Secure Connect provides encrypted transport. Haystack provides semantic tagging; Brick and ASHRAE 223P provide full semantic models. MQTT brokers with IAM frameworks manage publish-subscribe access. Cloud platforms provide service mesh and identity management. Zero-trust overlays like SPIFFE/SPIRE bind identity to workloads. Each of these addresses a real problem—but none of them governs the relationship itself.
Transport security ensures data is encrypted but does not define what a valid connection is. Semantic tagging and models provide shared vocabulary and structured representations but do not enforce contractual obligations between systems. Publish-subscribe brokers manage message routing but treat authorization as an access control list, not as a contextual, role-based contract. Cloud IAM and service mesh models can verify who is connecting but not why, in what role, or under what terms.
CNS/CP operates in the space that none of these tools occupy: the governance of the relationship between systems. It is complementary to all of them. The binding layer does not replace existing infrastructure; it makes existing infrastructure trustworthy by adding the governed relationship that is currently missing.
In Part 2, we’ll look at how CNS/CP actually works—the core architecture, a step-by-step walkthrough of a governed connection, and what the PAE project looks like through this lens.
References: From Models to Meaning (AutomatedBuildings.com) | When Platforms Stop Fighting (AutomatedBuildings.com)
Links: Project Arete | CNS/CP Specification | Arete SDK
