Documentation, data & digital twins
Every model, simulation and panel ends as data — fabrication files, drawings, schedules and an as-built record that, kept alive, becomes the digital twin running the facade for forty years.
The facade ends its design life as a number in a schedule — and begins its real life as that same number on a maintenance tablet, forty years later.
A facade's design is over the day the last panel is drawn, but the facade's data is just getting started. Everything upstream — the BIM model, the parametric geometry, the simulations — converges into a single deliverable: data. Data that drives the CNC machines cutting aluminium and glass. Data on the drawings and schedules the contractor builds from. Data on the as-built record. And, increasingly, data that lives on after handover as a digital twin — a maintained model of the real facade that the building's operators use to clean it, repair it, swap a cracked pane, and one day reclad it. The discipline of facade engineering does not end at installation; it ends, if it ends at all, when the building is demolished. This lesson follows the data the whole way.
From model to fabrication to a living asset record
The model becomes fabrication data — drawings, schedules and the files that drive the machines
A facade model's first job after design is to make the facade fabricable. The LOD 400 contractor's model generates three things. Shop/fabrication drawings: the detailed, dimensioned drawings of every profile, gasket, bracket and panel — still the legal and practical instrument the workshop builds from. Schedules: the panel schedule, glass schedule, bracket schedule — automatically extracted from the model, so a mullion's length appears once in the geometry and flows into every count and cut-list without re-typing.
The third is the powerful one: direct-to-machine fabrication data. Cut coordinates, drilling positions and panel geometry export straight from the model to CNC cutting, drilling and routing — closing the loop from design to manufacture with no manual re-drawing, which is exactly how thousands of unique parametric panels become makeable.
This is the spine again: because the facade was modelled as a coordinated system, one change to a panel updates its drawing, its schedule and its cut-file together. Break that link — re-draw a schedule by hand — and you reintroduce the drift BIM existed to kill.
A number typed once and flowing from model to drawing to schedule to CNC is BIM working. The same number re-typed in three places is where errors breed.
A facade is also a structured data set — what every panel and system carries, beyond its shape
Beyond geometry, a well-built facade model is a structured data set: every panel and system carries the information someone will need later. For each panel: its mark/ID, position, glass make-up and IGU spec, frame system, fire rating, U-value and SHGC, supplier, batch and date, warranty terms. For the system: the performance spec it was tested to, the mock-up reports, the maintenance regime.
This is the 'I' in BIM finally paying off downstream. A facility manager who wants to replace a cracked pane on the 24th floor should be able to query the model, find that exact panel's IGU specification and supplier, and order a matching replacement — instead of hiring someone to abseil down and measure it. A facade auditor checking embodied carbon should find the materials and quantities already in the data.
Under ISO 19650, this hand-over data set is formalised as the Asset Information Model (AIM) — the structured information about the built asset, handed from the project team to the operator. The honest caveat: the AIM is only useful if it is accurate at handover and maintained afterwards. A data set that was never reconciled to the as-built reality, or that froze on handover day and drifted ever since, is a liability dressed as an asset.
Kept alive and connected, the facade data set becomes a digital twin across the whole lifecycle
A digital twin is the as-built data set kept live — a maintained digital model of the real facade, ideally connected to real data (inspection records, sensor readings, repair history), used to operate and maintain the asset across its life. It is the difference between a model that was true on handover day and a model that stays true.
For a facade specifically, the twin supports the long tail the rest of this course set up: maintenance and cleaning (which panels, what access, the BMU routes from Module 8), repair and replacement (find a failed panel's exact spec and source a match), performance monitoring (is the facade leaking, thermally bridging, degrading?), and ultimately end-of-life — recladding or disassembly, where knowing exactly what is on the building and how it comes apart is the start of circularity (Module 11).
The honest reality as of 2026: full, sensor-connected digital twins of facades are still rare and mostly on flagship buildings; most projects, in India and globally, hand over a static as-built model and a document set rather than a living twin. But the direction is clear and the value is real — the facade is a forty-year asset, and the data that designs it is the same data that should run it. The engineer who treats handover data as a tick-box, not as the seed of the asset's operating life, throws away most of what BIM was for.
Think past handover. The data you specify for the facade model is what the building's operators will live with for decades — so ask, at the brief, what information the facility manager will need to clean, repair and eventually reclad this skin, and require it in the model. A beautiful as-built model with no panel IDs, no glass specs and no supplier data is a render, not an asset. Champion the digital twin where the project can sustain one, but be honest that a well-structured, accurate static data set well maintained beats an ambitious twin that no one keeps current.
You define and verify the facade data set. Decide what data each panel and system must carry (ID, spec, performance, supplier, warranty), set it in the BEP's information requirements, and — critically — verify it is ACCURATE at handover, reconciled to the as-built. Own the link from model to fabrication: schedules and CNC data extracted from the model, never re-typed, so a change propagates everywhere. Where a digital twin is in scope, define what it connects to and who maintains it; an unmaintained twin is worse than an honest static record, because it lies with confidence.
On site you help create the data the twin depends on. Every panel installed against its mark number, every survey of where it actually landed, every deviation recorded — that is the raw material of the as-built model. When you fit a panel, you are not just hanging glass; you are confirming (or correcting) a data record that someone will rely on in twenty years to find and replace it. Accurate as-built data is a site responsibility as real as a watertight seal — a panel logged wrong is a panel no one can find when it fails.
ISO 19650-3
Operational-phase information (AIM)
Governs information management in the operational phase — the Asset Information Model handed to the operator. It defines the PROCESS for asset data, not its facade content; and it only delivers value if the data is accurate at handover and maintained, which it frequently is not.
IFC / COBie
Asset data exchange & handover
IFC carries the model; COBie is a structured spreadsheet-style schedule of asset data (equipment, components, spaces) for facility management. Useful for handing over facade component data, but COBie can become a vast unmaintained spreadsheet if no one owns it after day one.
ISO 23247
Digital twin framework (manufacturing-origin)
A framework for digital twins of physical assets. Emerging in construction rather than established; as of 2026 full facade digital twins are rare and mostly on flagship projects, in India and globally — the direction is clear, the routine practice is not yet there.
NBC 2016 (India)
Documentation context
Indian projects still hand over conventional documentation under NBC and local bye-laws; structured BIM asset data and digital twins are an additional, project-driven layer on top, common on premium work but not mandated nationally as of 2026.
“Once the building is handed over, the facade model and its data have done their job and can be archived.”
Handover is where the data's most valuable life begins, not where it ends. A facade is a forty-year asset that will be cleaned, repaired, monitored and eventually reclad — and every one of those tasks is cheaper and safer when the as-built data set (panel IDs, specs, suppliers, performance) is accurate and maintained. Archiving the model on handover day discards most of what BIM was for. The data should live as long as the facade does, kept current as a digital twin or at least an accurate, accessible asset record.
Worked example — build a panel data record and test the replacement query
The test of a facade data set is whether a facility manager can find and replace one panel in twenty years without a survey. Let's build the data record for a single panel and run that query — the practical proof that the data is an asset, not a decoration.
Pen and paper, or a spreadsheet. No specialist software — the discipline is in the data, not the tool.
Define the data record one facade panel MUST carry to be findable and replaceable later: PANEL ID : ___ (e.g. P-24-07, floor 24, bay 07) POSITION : grid ref / level GLASS MAKE-UP : ___ (e.g. 6 Low-E / 16 Ar / 6 mm IGU) U-VALUE / SHGC : ___ / ___ FRAME SYSTEM : ___ (supplier system code) SUPPLIER/BATCH : ___ / ___ / date WARRANTY : ___ years from ___ TEST: in 2046 a pane cracks. Can you order a match from this record ALONE, without sending anyone up?
- 1Fill the PANEL ID and POSITION: a unique mark (P-24-07) tied to a grid reference. Without a unique ID, no later query can resolve to ONE panel — this is the key the whole record hangs on.
- 2Record the GLASS MAKE-UP and performance: the full IGU build-up (glass / cavity / glass, coating) plus U-value and SHGC. This is what a replacement must MATCH — get it wrong and the new pane looks or performs differently from its neighbours.
- 3Record the frame system and SUPPLIER/BATCH/date: the system code and who made it, when. In 2046 this is how you find whether the original supplier (or an equivalent) can still supply a match.
- 4Run the replacement query: in 2046 P-24-07 cracks. From this record alone you know its exact IGU spec, frame, supplier and position — you can order a matching pane and brief the access (BMU route) WITHOUT sending a surveyor up the building. That avoided survey is the data set paying off.
- 5Now test the failure mode: strip out the supplier and glass make-up (the 'render' version of the model) and re-run the query — you cannot order a match, so someone must abseil down to measure and identify it. The DIFFERENCE between those two outcomes is exactly the value of a real facade data set over a pretty model.
You’ll walk away with
A single-panel data record and a replacement-query test — the concrete demonstration of why a facade is a structured data set, not just a 3D model, and what the digital twin is ultimately for.
A closing reflection for the module.
- 01Picture a real glass tower near you with a single visibly cracked or fogged pane. Ask: does anyone know that exact panel's specification and supplier — or would replacing it need a survey? That gap is the difference between a facade with a living data set and one without; it is what this whole module set out to close.
Every model, simulation and parametric panel converges into data: fabrication files that drive the CNC machines, drawings and schedules extracted (never re-typed) from the model, and a structured facade data set carrying each panel's ID, spec, performance and supplier. Kept accurate at handover and maintained afterwards — as an Asset Information Model and ideally a digital twin — that data runs the facade across its whole forty-year life. The facade does not end at installation; its data should live as long as the skin does.
The model produces fabrication data: shop drawings, schedules (extracted, not re-typed) and direct-to-CNC cut files. A facade is also a structured data set — per-panel ID, glass make-up, U-value/SHGC, supplier, warranty — formalised at handover as the Asset Information Model (ISO 19650-3). Kept live as a digital twin, it supports maintenance, repair, monitoring and end-of-life recladding. As of 2026 full facade twins are rare; most projects hand over a static as-built record — accurate and maintained is what matters.
What is a digital twin of a facade?
A digital twin is the as-built facade data set kept live — a maintained digital model of the real facade, ideally connected to inspection, sensor and repair data, used to operate and maintain the skin across its whole life. It supports cleaning and access planning, finding and replacing failed panels, performance monitoring, and end-of-life recladding. The difference from an ordinary as-built model is that a twin stays current with the real building rather than freezing on handover day.
What data should a facade model carry for each panel?
Each panel should carry a unique mark/ID and position, its glass make-up and IGU specification, U-value and SHGC, the frame system, the supplier, batch and date, and warranty terms; the system level adds the performance spec, mock-up reports and maintenance regime. This is the structured 'I' in BIM that lets a facility manager later query the model to find and order a matching replacement for a failed panel without sending a surveyor up the building.
Are facade digital twins used in India?
Rarely, as of 2026, and mostly on flagship projects — in India and globally alike. Most Indian projects hand over a conventional document set and at best a static as-built BIM model rather than a live, sensor-connected digital twin. The direction of travel is clear and the value is real for a forty-year asset, but routine practice still relies on as-built models and asset schedules. What matters most is that the handover data is accurate and maintained, not whether it is branded a twin.
Peer-reviewed journals & authoritative standards
- 01Su, Z. et al. Multi-Disciplinary Characteristics of Double-Skin Facades for Computational Modeling Perspective and Practical Design Considerations. Buildings, 12(10):1576. — Buildings (MDPI), 2022.
- 02Review on Glass Curtain Walls under Different Dynamic Mechanical Loads: Regulations, Experimental Methods and Numerical Tools. IntechOpen. — IntechOpen, 2023.
- 03Material Selection and Characterization for a Novel Frame-Integrated Curtain Wall. (PMC8069006). — Materials / NCBI-PMC, 2021.
That completes the digital workflow — model, computation, simulation and data carrying the facade from concept to a living asset. The final module turns to where all this data points: the facade's future, where the asset record becomes the starting point for embodied-carbon accounting, recladding and the circular, low-carbon skin.