TL;DR:

  • Museums are deploying AR to overlay context on physical objects and XR to reconstruct spaces that no longer exist
  • The most successful implementations add information that genuinely can’t be conveyed with traditional labels — not just digital versions of existing content
  • Accessibility and remote access are emerging as the strongest justifications for XR investment, beyond on-site visitor enhancement

When the British Museum deployed an AR layer on selected objects in 2024 — letting visitors point a tablet at a mummy case to see the person as they looked in life, based on forensic reconstruction — it demonstrated what museums are increasingly recognising: spatial computing offers capabilities that no previous medium could provide. You can’t put a person’s reconstructed face on a label. You can’t animate the mechanisms of a clock on a placard. But you can in AR.

This has pushed cultural institutions that previously viewed XR as a gimmick into serious exploration of where the technology genuinely enhances their core mission: connecting people with history, art, and culture.

Where XR Is Delivering Real Value in Cultural Institutions

Archaeological Site Reconstruction

This is perhaps the strongest use case in heritage, and the reason institutions like the Smithsonian, the Acropolis Museum in Athens, and Historic England have invested in it seriously.

When a physical site is ruined, fragmentary, or inaccessible, XR reconstruction allows visitors to see it as it was. The Acropolis Museum uses VR stations that let visitors walk through the Acropolis as it appeared in 440 BCE — the Parthenon intact, painted in the colours that analysis of trace pigments reveals, with the cult statue of Athena inside. This is information that simply cannot be communicated by looking at the surviving broken columns.

The technical pipeline for these reconstructions has improved substantially. Photogrammetry from drones and ground-level cameras creates detailed 3D models of surviving remains. Art historians and archaeologists provide the interpretive layer — deciding where evidence supports reconstruction versus where the model is speculative. The result can be rendered in a standalone headset (Meta Quest 3 is the most common deployment choice for museums due to its price point and ease of management) or as a browser-based WebXR experience that visitors access on their own devices.

Object Context and Hidden Information

Physical objects in museums are often more interesting beneath or behind what’s visible. X-ray analysis of a painting reveals earlier compositions underneath. CT scans of sealed Egyptian artefacts reveal their contents without opening them. Infrared imaging shows preliminary sketches beneath the final paint layer.

AR makes all of this accessible without any physical intervention. A visitor holding a tablet or wearing AR glasses in front of a specific painting can see the underdrawing as an overlay, or watch the painting as it appeared when first completed versus its current degraded state.

The Rijksmuseum in Amsterdam pioneered this approach with Rembrandt’s Night Watch during its conservation project — broadcasting the conservation work live and creating AR experiences that showed restoration in progress. The combination of physical presence and digital enhancement reached audiences that neither the physical object alone nor a digital-only experience would have.

Making Inaccessible Spaces Accessible

Some of the most significant cultural heritage sites are physically inaccessible — either because they’re too fragile to open to visitors, geographically remote, or exist only in historical records.

The caves at Lascaux in France closed to tourists in 1963 because CO2 from visitors’ breath was damaging the 17,000-year-old paintings. Lascaux IV, opened in 2016, includes physical replicas, but VR experiences can take visitors inside a faithful reconstruction of the original cave in a way that the replica cannot fully replicate.

Similarly, sites damaged or destroyed by conflict — Palmyra in Syria, the Buddhas of Bamiyan — exist in detailed 3D documentation that XR experiences can bring to audiences who will never see the physical originals.

Remote and Inclusive Access

Museums are increasingly recognising that their primary audience constraint is physical geography. A school in rural Scotland cannot easily bring students to the National Museum in Edinburgh, let alone the Louvre. XR provides access that doesn’t require the journey.

The Natural History Museum’s virtual field trips — available as browser-based experiences accessible from any device — let school classes explore the collections as a structured educational activity. The engagement data from these programmes shows longer session times and higher recall scores than equivalent video-based tours, which is driving continued investment.

For visitors with mobility impairments, XR enables access to areas that physical accessibility limitations make difficult — staircases, cramped historic spaces, outdoor sites with uneven terrain.

Implementation Choices: What’s Actually Working

Device strategy: Shared headsets vs. personal devices

Most institutions offering on-site VR experiences use a pool of managed headsets — typically Meta Quest 3 at the £500 price point — rather than expecting visitors to bring their own. Headset hygiene (cleaning headbands and face gaskets between visitors), battery management, and queue management at popular experiences are the operational challenges.

AR experiences accessed on visitors’ own smartphones are simpler to operate — no hardware management — but require clearly marked markers or NFC triggers to initiate, and the experience quality on a phone screen is lower than through a headset.

Apple Vision Pro has been trialled at a few institutions for premium guided experiences where display quality matters for art — the Tate Modern used it for a small-group immersive tour of their modern painting collection. The cost limits it to selective use.

WebXR for remote access

For remote and educational access, WebXR — XR experiences that run in a browser without an app download — is the dominant approach. Google’s model-viewer component, A-Frame, and Three.js enable museum 3D object viewers that work on any device. Sketchfab hosts thousands of digitised museum objects in explorable 3D.

The British Museum, Metropolitan Museum of Art, and several European national museums have made significant collections available as WebXR-compatible 3D models, licensed for educational use.

Curatorial standards for reconstruction

Institutions that have done this well are clear about epistemic status: what is documented fact versus scholarly reconstruction versus artistic interpretation. Good XR heritage experiences label these distinctions visibly — “this reconstruction is based on fragmentary evidence and represents one scholarly interpretation” — rather than presenting speculation as certainty. This is both an intellectual integrity matter and a protection against criticism when the reconstruction diverges from future discoveries.

The ROI Question

For public institutions operating on constrained budgets, the justification for XR investment needs to be clearer than “innovative visitor experience.” The strongest cases being made are:

  • Visitor dwell time and return visits: Institutions with compelling XR experiences see longer average visits and higher rates of return visits, which drives membership and repeat admission revenue.
  • Educational programme revenue: School group bookings tied to curriculum-linked XR experiences are a growing revenue stream.
  • Remote access: Digitisation and XR programmes that reach global audiences online build the institution’s profile and drive donations from international supporters who may never visit in person.
  • Conservation: Digital documentation via photogrammetry and 3D scanning — the same data used for XR experiences — has independent conservation value, making the content creation investment dual-purpose.

The institutions making the most progress are treating XR as a content production and distribution strategy, not as a technology experiment.