MondayLive! is spending May unpacking what foundational readiness looks like in an era of accelerating change. Early sessions asked each participant what “basic” meant to them personally, ranging from mechanical fundamentals to business operations. From there, the conversation evolved into a harder question: what is the new basic, given that the industry is in the middle of a genuine transition?
This session focused specifically on Basic Readiness, which can be understood as the minimum conditions a building must meet before any meaningful AI-driven analysis, optimization, or operation can take place.
The news and trends segment set the stage well. There is sustained, broad interest in IoT protocols and their integration with knowledge graphs. BACnet troubleshooting content continues to draw heavy traffic and renewed attention at AutomatedBuildings.com. The concept of renaming a Building Management System to a “Building Operating System” is gaining traction, reflecting a shift from passive monitoring toward active operations. APIs, long considered table stakes, are being rediscovered as the connective tissue that will link legacy systems to modern platforms.
A recurring theme across all of it: we are the actuators of AI. Sensors, controls, field devices, and the people who commission and maintain them. Without that physical and operational layer working correctly, AI has nothing real to act on.
AI Is Already Commissioning Buildings
One concrete signal of where things are heading came from a report on the Controls-Con event held in Denver, sponsored by Cochrane. The conference theme was AI, and one standout demonstration involved a commissioning application that uses AI to ingest point data from a newly installed system and determine, over a period of days, whether the system is operating per specification.
Rather than requiring a field engineer to manually verify each point, the application performs the analysis automatically and flags deviations. For firms that spend significant time proving out system compliance before handoff, this represents a meaningful reduction in labor and risk.
Commissioning kept coming up throughout the session as one of the cleaner early applications of AI in buildings, precisely because the task is well-defined. You have a spec. You have sensor data. The question of whether they match is answerable. A dedicated Monday session on commissioning was proposed as part of the readiness conversation.
Separately, a panel discussion this week at the US Green Building Council, part of a webinar series titled “Aligning Priorities for AI in Existing Buildings: From Hype to What Matters,” reflects how broadly the AI question has spread. Energy teams, operations teams, engineering firms, and sustainability practitioners are all arriving at the same intersection from different directions.
The Foundational Problem: You Don’t Know Where Anything Is
The core of the session was a frank look at the state of building documentation, and it was not flattering.
The conversation opened with a point that sounds almost too simple to say out loud: most buildings do not have a consistent, agreed-upon room numbering system. Not a shared digital model. Not even consistent room numbers. A floor plan from one contractor does not match the records in the facilities management system, which does not match what was entered into the work order platform three tenants ago.
“If you can’t point to a map and say where your system is, then it’s just a signal without relationship.”
The downstream consequences of this are significant. If you do not know where a space is, you cannot meaningfully claim to know what energy that space consumes. You cannot verify whether a 30 percent savings figure applies to a critical zone or a storage closet. You cannot connect a sensor reading to the equipment it belongs to, the room it serves, or the people affected by its performance.
This is the current state of the art in most buildings. And it is why promising AI-driven outcomes on top of that foundation produce numbers without context rather than intelligence.
The Building Readiness Stack
The group worked through a layered model of what readiness actually requires, building up from the most foundational element. The graphic below captures that structure.
Level 0: Context and Geospatial Location
Before anything else, a building must exist as a known, located object. A pin on a map. An agreed identity. This sounds trivially obvious until you learn that one organization discovered it had two different database entries for the same building because no one had verified geospatial coordinates across systems. When two pins land in the same spot, that is immediately visible and resolvable. Without the pin, the duplication is invisible.
This level was proposed during the session as a precondition for everything else, a Level Zero that sits beneath even documentation. The building has to be somewhere before it can be anything.
Level 1: Documentation
Floor plans. As-built drawings. Design specifications that include intended airflow, load capacity, and energy targets. The documents do not need to be perfect. They do not need to be digital twins with 3D fly-through capability. A 2D floor plan, even a scanned one, can be ingested by AI today and converted into a basic spatial model in minutes.
The key insight here is the distinction between basic and elaborate. The industry has been sold on cinematic digital twins that are expensive to build and impossible to maintain. The actual need is far simpler: a document that establishes what a building contains, where those things are, and what they were designed to do. That is the crawl stage. Most buildings are not crawling.
Level 2: Organization and Source of Truth
Once documentation exists, it must be organized around a single authoritative identifier system. Room numbers. Asset IDs. A list that the owner owns, maintains, and distributes to every contractor, vendor, and system that touches the building going forward.
The California Community Colleges system was cited as a rare example of an owner doing this at scale. Roughly 90 million square feet of facilities are maintained with a consistent space inventory, because the funding mechanism makes it mandatory. No accurate inventory, no bond money for construction or renovation. The incentive structure enforces the behavior.
Without that kind of enforcement, fragmentation is the default. Every new project brings new drawings with new room numbers. Every system migration imports the chaos from the last one. The work order platform becomes the de facto asset list, not because it was designed for that purpose, but because it is the system people actually use. And even that drifts without active maintenance.
The standard at the project level for this is COBie, a building smart specification that defines how spaces and assets should be listed with unique IDs. It can be written into architectural contracts. The barrier is not technical. It is that most owners do not know to ask for it.
Level 3: Machine-Readable and Open Formats
Data that exists only in a proprietary system is not really accessible. It is held. The difference matters when you need multiple systems to share information, when you need AI to analyze it, or when you need to migrate away from a vendor that no longer meets your needs.
Open formats, specifically RDF turtle files, IFC, and common spreadsheet formats, allow data to move between systems without requiring a license, a custom integration, or a consultant to translate. A six-day experiment with the PAE Living Building demonstrated this concretely. A turtle file representing the building was shared with an external partner, who added 100 new elements maintaining the existing IDs throughout, and returned the file. No meetings. No API negotiation. Just a shared file format and agreed ID conventions. The transaction was machine-to-machine.
The files from that project, including turtle files, Excel exports, a Revit model, a cloud BIM version, and an API, are being posted to GitHub as the work progresses, with the intent of giving the industry a concrete reference point rather than an abstract recommendation.
Most major platforms can export in open formats. They do not advertise this capability. Requiring it in contracts is the lever.
Level 4: Exportable and Connectable
A periodic export is a snapshot. A snapshot goes stale. The distinction raised in the session is that data must be both exportable, available as a file at any point, and connectable, accessible live through an API, so that systems can stay synchronized without waiting for the next scheduled dump.
“Exportable” gives you a moment in time. “Connectable” gives you the truth, ongoing. Both are required. A system that can only export is a silo that occasionally opens its doors. A system that exposes a live connection over an open protocol becomes a participant in a broader information ecosystem.
The Accounting Analogy
One of the sharpest framings of the session drew a direct comparison to business accounting. No organization of any size operates without an accounting system. Income, expenses, cost of goods, categories, reconciliation. These are maintained continuously and taken seriously because the consequences of not doing so are immediate and tangible: you cannot pay salaries, file taxes, or understand whether the business is viable.
Buildings do not have an equivalent. There is no system that building owners are culturally or contractually expected to maintain with the same rigor. The closest candidates are IWMS and CMMS platforms, and some organizations use them well. But the norm is that documentation is created during construction and then ignored, updated inconsistently, or lost entirely as tenants turn over, systems get replaced, and floors get rebuilt.
“That attitude does not exist in our industry. There is no analogous system that provides the necessary documentation in an organized fashion that is maintained, if you will, religiously.”
The proposal that followed was not to build a new system but to treat the existing data, whatever it is and wherever it lives, as the starting point. A pencil-written list of room numbers, photographed and run through AI to extract the text, becomes machine-readable data. That is enough to start. What matters is that the owner takes ownership of that output, parks it in an accessible place, and requires every future contractor to reference it and return to it.
Contractual Requirements Are the Mechanism
Several practical paths toward enforcing these standards came up in the discussion. The most direct is contract language. For projects that use an AIA contract or similar instrument, requirements for space lists, asset IDs, and open file formats can be written explicitly. The State Department was cited as an example of an owner that has done this, requiring delivered assets to carry IDs that match a defined standard and verifying compliance during the project rather than at handoff.
The broader framework for this kind of standardization is being developed through the Coalition for Smarter Buildings (C4SB), where a Digital Building Profiles working group is defining what information should be known about any building, in what format, and at what level of detail. The effort is intended to produce something concrete enough to be referenced in construction specifications, specifically the Division 25 section that covers integrated automation, so that engineers and owners have a shared vocabulary and a clear deliverable to request.
The town hall recording from C4SB’s recent gathering is available on the C4SB website and covers the status of multiple working groups, including Digital Building Profiles.
Crawl First
The session closed with a summation of what May’s “Back to Basics” focus had produced. The group did not reach a final framework. What they arrived at was a clearer picture of where the gaps are and why filling them does not require waiting for the next major platform, the next construction project, or the next budget cycle.
Basic readiness work can be done during a maintenance visit. It can happen when a technician is already on-site and adds a room number to a record. It can happen when an owner changes work order systems and uses the migration as an opportunity to clean the data rather than carry the chaos forward.
The argument being made is not that perfection is required before progress is possible. It is that the direction matters. A building moving from no machine-readable data toward a single agreed-upon room list is moving in the right direction. A building adding sensors and AI dashboards on top of undocumented, conflicting records is not.
“The basics are basic. They are not that difficult to do. Unless you try it. And as soon as you try it, then you see it.”
There will be no MondayLive! next week due to the Memorial Day holiday. The session resumes in two weeks. Slide decks from each session are available at mondaylive.org, and views expressed are personal, not those of any company or organization.
