Create a Digital Twin Of Your Physical Model
A physical architecture model communicates qualities that drawings can struggle to show: the grain of a material, the thickness of a wall, the relationship between a courtyard and its surroundings, or the way a building sits within a larger urban fabric. An online visitor cannot walk around the table, change their viewing height or lean closer to inspect a detail. A digital twin can bridge that gap by creating a navigable digital counterpart of the exhibited object.
For a student project, the twin does not need to be a perfect engineering replica. It needs to be a reliable, clearly labelled and accessible representation of the physical model, with enough visual information to help viewers understand the design. This might be a web-based 3D viewer, a set of connected panoramas, an augmented-reality experience or a lightweight interactive scene.
The strongest results come from treating the online version as part of the exhibition rather than as a last-minute scan. Decide what visitors should learn, select a capture method that suits the model’s materials, and write explanations that add meaning without competing with the design. A carefully prepared digital presentation can support jurors, classmates, educators and members of the public who cannot attend in person.
This approach is especially useful for exhibitions that reach visitors across Australia and beyond. Someone viewing a project from Melbourne, Adelaide or regional New South Wales may have a different screen, internet connection and amount of time available than a visitor standing in a gallery. The digital twin should therefore be informative at a glance, detailed when explored and efficient to load.
Define The Twin Before You Capture It
Begin by defining the purpose of the online model. A competition jury may need to understand circulation, structure and the relationship between drawings and form. A general visitor may first want to know the project’s location, scale and social purpose. A prospective student may be more interested in the making process, including failed studies, material tests and changes made during the semester.
Write a short viewing brief before opening a camera or modelling program. Identify three to five ideas that must survive the move from physical exhibition to screen. For example, a housing proposal might prioritise the shared threshold, the daylight strategy and the relationship between private rooms and a communal garden. Every scan, annotation and image should support at least one of these ideas.
Choose the appropriate level of detail. A detailed mesh may preserve tiny surface features, but it can become slow on a phone and difficult to edit. A simplified model can load quickly and offer a cleaner overview, while close-up photographs provide evidence of texture. Many successful presentations use both: a lightweight navigable object for orientation and a small gallery of high-resolution images for inspection.
Record essential metadata alongside the geometry. Include the project title, student or team name, institution, year, scale, materials and software used. If the original object is 1:100, explain how a visitor should interpret its dimensions. Clear information prevents viewers from mistaking a conceptual massing model for a construction-ready digital model.
Capture The Physical Model Accurately
Photogrammetry is often the most accessible way to produce a digital model. Place the object on a stable surface, photograph it from overlapping angles and process the images in software that can generate a mesh and texture. Keep the camera at a consistent distance, use diffuse light and avoid strong shadows that may be interpreted as permanent features. A turntable can help with small objects, although a fixed camera and carefully marked positions can work just as well.
Scan the model in a controlled environment. Australian daylight can be intense, particularly near north-facing windows in Brisbane, Sydney or Perth, so direct sun may create changing highlights between photographs. Use soft, even lighting and keep the model clean. Glossy acrylic, transparent film, mirrored surfaces and very thin elements are difficult for photogrammetry because the software cannot easily identify stable points. Photograph these components separately and reconstruct them manually if required.
For larger or highly precise models, consider structured-light scanning or LiDAR. A recent phone or tablet may provide useful depth information for broad forms, but the result can lose fine edges and dark surfaces. A professional scanner offers more control, yet the cost may not suit a student exhibition. Combining a modest scan with manual modelling in Blender, Rhino, SketchUp or Revit can produce a better outcome than relying on a single capture method.
Use scale references and quality checks. Place a ruler or known-size marker near the model during capture, then verify important dimensions after processing. Compare the digital twin with the physical object from several familiar viewpoints. Check roof edges, openings, slender columns, stairs and landscape elements, because these are common areas where software fills gaps or smooths away meaningful features.
Document the making process as well. A photograph of the model beside its screen-based counterpart helps online visitors understand the relationship between the two. Process images can also explain where the digital version has been simplified. That honesty is valuable: the twin is an interpretive exhibition tool, not a claim that every physical fibre or handmade imperfection has been reproduced.
Prepare A Web-Friendly 3D Asset
Before uploading, clean the captured mesh. Remove hidden faces, isolated fragments and accidental background geometry. Close holes where a visitor might see through the object, but do not automatically smooth every irregularity. Handmade cuts, rough edges and material changes can communicate the project’s character. Retain them when they support the design, and simplify them when they only increase file size.
Create several levels of detail, often called LOD versions. The first version should open quickly and show the entire project. A second version can include more geometry for visitors who choose to inspect it closely. Formats such as GLB or glTF are commonly suited to browser-based 3D because geometry, materials and textures can be packaged efficiently. Compress textures, use sensible image dimensions and test the result on a mid-range mobile device rather than only on a powerful studio computer.
Optimise the page around the model. Display a still image or lightweight preview before loading the interactive viewer, and show a visible loading state so visitors know that progress is occurring. Provide orbit, pan and zoom controls with large touch targets. On smaller screens, a simple auto-rotate option can help visitors who are unfamiliar with 3D interfaces, but it should never replace manual control.
Plan hosting and file delivery early. Australian visitors may use fast NBN connections at home, but mobile networks and regional internet services can be less consistent. A large uncompressed file can discourage access even when the model itself is compelling. Use responsive images, browser caching and a content delivery network where appropriate. Test from a standard home connection and a phone using mobile data.
The digital exhibition can sit beside a concise project statement and downloadable drawings. A visitor who wants a quick overview should not need to operate the 3D viewer to discover the project’s premise. For examples of how student work can be presented within an architectural exhibition archive, explore the past project works and note how titles, images and supporting information guide attention.
Build An Experience For Real Visitors
A digital twin becomes useful when it tells a spatial story. Set an initial camera position that shows the overall composition, then offer suggested viewpoints for the entrance, principal interior, public realm and important detail. These positions can be simple buttons or labelled hotspots. Avoid forcing every visitor through a long animation; jurors and experienced designers may prefer to move freely.
Annotations should explain decisions rather than describe what is already visible. “Recycled brick screen filters afternoon sun” is more helpful than “brick wall”. Keep labels short, place them away from busy geometry and ensure they remain readable on a phone. A separate text panel can provide longer notes about climate, structure, material sourcing or the social conditions addressed by the proposal.
Accessibility must be designed from the beginning. Provide a keyboard-accessible interface, descriptive alternative text for still images and a text-based summary of the spatial sequence. Do not communicate meaning through colour alone, and give users a way to pause rotation or animation. These practices align with the Web Content Accessibility Guidelines and support obligations associated with Australia’s Disability Discrimination Act 1992, while also helping visitors using screen magnification, older devices or noisy public environments.
Remember that the exhibition may be viewed at very different times. If a live opening or online talk is scheduled across Australia, state the time zone and distinguish AEST from AEDT when daylight saving applies. A visitor in Perth, for example, may need to convert a Melbourne event time before joining. Recorded walkthroughs, captions and a stable project page make the work available beyond the original programme.
The exhibitor information should also be easy to find. When preparing submission details, review the exhibitor resources so the digital twin complements required images, statements and file formats rather than duplicating them. A consistent presentation helps a project sit comfortably beside physical boards, photographs and other student entries.
Publish Responsibly And Keep It Useful
Before publication, run a practical review with someone who did not make the model. Ask them to find the main entrance, identify the project’s scale and explain its central idea after a few minutes. Their confusion will reveal problems that the modelling team may overlook because they already know every decision. Test the page in Chrome, Safari and a common mobile browser, and check portrait and landscape orientations.
Protect people and information connected with the project. Do not include identifiable faces, private addresses or studio documents in the scan unless you have permission. If the page collects analytics, newsletter details or comments, explain what is collected and why. Australian organisations handling personal information may need to consider the Privacy Act 1988 and the Australian Privacy Principles, especially when a platform stores data overseas or shares it with third-party services.
Check intellectual-property permissions for photographs, textures, music, fonts and software assets. A student may own the model while a collaborator owns a photograph or a studio retains rights in a supplied material library. Keep an internal record of permissions and attribution. Also provide a clear credit line for the student, institution, photographer, modeller and any technical assistance.
Use the following publication checklist before the link is shared:
- Verify the digital twin against the physical model from overall and close-up viewpoints.
- Export a compressed web version and test loading on a phone using mobile data.
- Add a project summary, scale, materials, author details and image credits.
- Provide keyboard controls, alternative text, captions and a text-only spatial description.
- Check that no personal information, unlicensed asset or confidential drawing is visible.
- Test links, download buttons, annotations and screen-reader labels.
- Save an archived master file so the exhibition can be updated without rescanning.
After launch, monitor the experience without changing the design simply to chase engagement statistics. Analytics can show whether pages load, where visitors leave and which devices are common, but a low interaction count does not necessarily mean that the project failed. Some visitors may read the statement, inspect two images and leave with a clear understanding. Preserve the original files, record the software versions and schedule a future migration if the viewer platform becomes obsolete.
A digital twin works best when it respects the physical model’s strengths while using the web for what the gallery cannot provide: multiple viewpoints, searchable information, captions, links and access across distance. The essential standard is simple: visitors should be able to understand the project’s scale, spatial idea and material character without needing the maker beside them. Build for that understanding, and the digital counterpart will remain a meaningful part of the exhibition rather than a decorative duplicate.