I believe we will build the first true home digital twin. Not a scan of its rooms, not a list of what’s in it, and not the builder’s files from the week it was finished. A counterpart of the whole house: every room, every system, every paper and every person who works on it, each fact showing where it came from. One you hold, open and check, and that gets truer every year you live there.

Every 3D house you’ve seen is finished the moment it loads
A scan shows the rooms as they looked on the day of the scan. A builder’s model shows the house the week it was handed over. A listing’s walk-through shows the house the week it went up for sale. Each is beautiful, each is drawn for one moment, and none of them can tell you where a single wall came from.
The companies that use the words “digital twin” for a home each hold one real slice of it:
- A scan. Matterport: “a dimensionally accurate 3D digital representation of a built space such as a residential or a commercial building.”
- The builder’s handoff. Digs turns a builder’s plans and documents into “a structured, intelligent digital twin of the home,” kept “from pre-con to final handoff.”
- A list. HomeZada: “a living digital record that represents the physical and financial details of your property.”
- A file for insurers. Cotality “builds digital twins of 178 million structures” from imagery, and says a twin that exists “before a loss” gives “a perfect record of the property’s condition.”
None of these definitions says how you tell a measured fact from an estimated one, and none puts the household in charge.
Where the idea comes from, and who it left out
Digital twins grew up in factories, plants and large buildings. The Digital Twin Consortium calls one “an integrated data-driven virtual representation of real-world entities and processes, with synchronized interaction at a specified frequency and fidelity.” In buildings, the twin starts at handover: COBie exists so a project team can “deliver the data necessary to maintain a facility in a digital format” to its owner.
In that world, the twin is geometry plus data plus the building’s whole life, and the data is the hard part. Homes were left out. Wikipedia’s article on digital twins doesn’t mention a home, a house or a homeowner. A 2022 review of digital twins for finding faults in buildings writes for facility managers and building owners; residents appear as occupants who “could fail to address the problem with facility managers.”
Consumer 3D starts from geometry and never gets the data. We start from the data, because it’s the part homeowners already have: the receipts, the manuals, the jobs, the people. Geometry attaches to it as it arrives.
The house that gets truer
On day zero, before you’ve told us anything, Homaic draws a house you recognize: your home’s real outline, laid out into the usual rooms by estimate, and it says the rooms are estimated. Where open map data has no outline for your home, Homaic draws no estimate, and the house starts from what you give it.
Then it develops, like a photograph. Every true thing you give it replaces an estimated part with a known one. Steps one to four work today, and the rest are coming:
- Estimated. The outline from open map data, the storeys, and the beds and baths from public records where they have them.
- Named. Your rooms replace the usual set, and your things stand in their rooms.
- Drawn. A plan you upload is fitted to the outline, and you draw your rooms on it.
- Scanned. A room scan puts the real walls, the door and the windows where they are.
- Placed and sized. The furnace stands where it stands, at its real size, from the manual or a tape measure.
- Systems. The panel map and the shut-offs, recorded, with what each one feeds.
- Finished. The floors and walls in the materials you recorded.
- Complete. Nothing estimated left. The house is the record, seen.
At every step the 3D house is a view of the record: a home digital twin built only from what’s true. The record stays the truth, and nothing lives in the model alone. That’s what lets it be honest on day zero and complete in year five, without ever being rebuilt.
A tap opens the record
In every 3D house I’ve seen, a tap ends at the object. Here, a tap on the water heater opens its record: who serviced it and when, what it cost, the warranty, and the page of the maker’s manual that says how to relight it. The model is a door into the record. Rendering is table stakes. The record under the room is the point.

What it means
For the worst night. Where does the water shut off, and what goes dark when breaker 14 trips? Direction, not a feature yet: tap the main valve or the breaker, and the house dims everything it feeds and keeps the rest lit, with a count. No pipes drawn through walls we haven’t seen. Just the consequence, from what’s recorded.
For insurance. A claim pays for what you can show you had. The Insurance Information Institute says an up-to-date home inventory helps you “get your insurance claim settled faster.” A twin is that inventory, room by room, with its receipts and photos, kept before anything goes wrong.
For selling the house. France requires a carnet d’information du logement for homes whose permit was filed from January 1, 2023. The UK government says it “will legislate to make both digital property logbooks and sales packs a requirement for all property transactions when parliamentary time allows.” A record that travels with the house is where both are going. How a home record maps to COBie, IFC and the logbooks.
For care. This works today. A thing’s care lines come from its manual, with the page they’re on, and one tap turns a line that says how often into a repeating task.
For the questions a list can’t answer. Will a 36-inch fridge fit? Where should the router go? What does taking out this wall touch? These need sizes with sources. Homaic will answer them with sizes that name their source, and say “I don’t know that size” when nobody recorded one. Today you can connect your own AI agent. It can read your record and add to it. It never deletes, shares, or changes your settings.
For the years ahead. Drag a day bar back and the house replays its own log: each job, paper and repair lands on its thing on the day it happened, and nothing between two records is drawn. Drag it forward and things age out one by one, from their install dates and the lifespans their makers publish, drawn as estimates. Direction, not a feature yet.
For your privacy. A house drawn accurately is sensitive. No link you share shows the 3D house. Later you’ll be able to switch it on for a single link. A scan’s photos stay private to your account.
Why this can’t be copied
Every company that draws your house today sells geometry for one transaction: a listing, a quote, a claim. Anyone can copy a 3D house in a weekend. Nobody can copy a house that has been kept true for years: the papers behind each fact, every change logged, every estimate replaced by something real.
What we will never do
- Call an estimate a measurement. A blank is a blank, not a guess.
- Let the picture become the record. Nothing lives in the 3D house alone.
- Control your devices from the model. Homaic records. It doesn’t run your house.
- Score your house. We say what’s known and what’s estimated, never a health grade.
- Pass off a generated picture as an observation. A drawing made by AI never counts as what’s there.
- Sell your data. We store it for you. We never sell it or use it for ads.
Where it stands
At launch: the estimated house on day zero, your rooms and things in it, plans fitted to the outline, room scans placed, and a tap from any thing to its record and its manual.
In the months after: what goes dark from the panel and the shut-offs, the house’s log replayed on the model, and sizes you can measure with their source.
In the year after: scans, sizes, install dates and a panel map on most houses, and the record in the shapes a sale asks for: AR, then the building-handover workbook, then IFC.
The full twin: every room its true shape, every system with what it feeds, every finish recorded, nothing estimated left. Each step comes when the record can carry it, and not before.
Sources (12)
- Matterport, Explore the Power of 3D with Digital Twins, May 28, 2020, read October 30, 2026
- Digs, Turning Construction Data Into a Living Digital Twin, May 12, 2026, read October 30, 2026
- HomeZada, Home Intelligence Platform, read October 30, 2026
- Cotality, Computer vision for property professionals, read October 30, 2026
- Cotality, How are digital twins reshaping property mitigation?, December 9, 2025, read October 30, 2026
- Digital Twin Consortium, The Definition of a Digital Twin, read October 30, 2026
- National Institute of Building Sciences, COBie, read October 30, 2026
- Wikipedia, Digital twin, revision of October 19, 2026, read October 30, 2026
- Hosamo et al., A review of the Digital Twin technology for fault detection in buildings, Frontiers in Built Environment, 2022, read October 30, 2026
- Insurance Information Institute, How to create a home inventory, read October 30, 2026
- Légifrance, Code de la construction et de l’habitation, L126-35-2 to L126-35-4, read October 30, 2026
- GOV.UK, Home buying and selling reform roadmap, updated June 19, 2026, read October 30, 2026