TL;DR: Museums, heritage sites and visitor attractions use XR to show what visitors cannot otherwise see: a ruin as it stood, an object from every angle, a story placed in the room where it happened. The format matters more than the technology. AR on visitors' own phones reaches the most people and suits outdoor sites and trails. Staffed VR stations suit a single, powerful experience inside a gallery. Mixed reality and shared sessions suit groups and school visits. The projects that work are scoped around visitor flow, staff capacity and the collection's own source material, and they plan for years of unattended use from the start. A good studio will also tell you when a well-made video or a printed panel would do the job better.
Cultural institutions were early adopters of immersive media, and many have the scars to show for it: a headset installation that impressed at the launch and sat switched off a year later, or an app that needed a download nobody bothered with. The technology is not usually what failed. The project was scoped as a showpiece rather than as a piece of visitor infrastructure that staff have to run every day.
This guide is for curators, digital leads, heritage trusts and attraction operators deciding whether an XR experience is worth commissioning, and what to ask before they do.
What XR is actually good at in a museum
XR earns its place when it shows something a visitor physically cannot see. That gives a useful test for any idea.
- Reconstruction. A ruined building, a lost street or a landscape as it looked centuries ago, shown where it stood or at the scale it had.
- Hidden views. The inside of a sealed object, the underside of a fragile piece that cannot be turned over, the layers beneath a painting.
- Things too big, small or fragile to display. Ships, aircraft, machinery, microscopic specimens, objects held in storage.
- Context and process. How an object was made or used, shown step by step in three dimensions rather than described on a label.
- Place-based storytelling. An event told in the location where it happened, layered over what is there today.
If the idea is mainly information that a panel, an audio guide or a short film could carry, XR adds cost and friction without adding meaning. That is worth saying in the first meeting.
Three formats, three different projects
"Museum XR" covers projects with very little in common once you scope them. Choose the format first, because it decides almost everything else.
AR on visitors' own phones
The visitor points their phone at an object, a gallery wall or a site and sees content placed on top: a reconstruction, an animation, a character, a label in their language. Nothing to hand out, clean or charge, and it can reach people outside the building, on a heritage trail or a city walk.
The trade-offs are reach versus friction. A native app can do more, but asking visitors to download one at the door loses many of them. WebAR that runs in the browser from a QR code removes that step, with some limits on tracking and visual quality. Outdoor sites bring their own problems: GPS accuracy, lighting, crowds walking through the scene and patchy signal. Our guide to location-based AR covers those in depth.
Staffed VR stations
A visitor puts on a headset and is taken somewhere: inside a tomb, aboard a ship, onto a battlefield, into a building that no longer exists. Nothing else matches the sense of being there, which is why VR is so often the first idea.
It is also the most demanding format to run. Every session needs someone to fit the headset, clean it between visitors, watch for discomfort and reset the experience. Throughput is limited by session length and the number of stations, so a twenty-minute experience in a busy gallery creates a queue. Comfort matters with a general audience that includes children, older visitors and people who have never worn a headset, so movement should be gentle or seated. VR motion sickness explains the design choices that keep people comfortable.
Shared mixed reality and group experiences
Several visitors, often a school group or a guided tour, see the same virtual content anchored in the same physical space and can talk about it together. A guide can bring up a reconstruction in the middle of a gallery and the group walks around it. This suits education programmes and premium tours, and it needs careful network and device planning. Mixed reality collaboration covers how shared sessions are built.
Start from visitor flow, not from the headset
The single most useful question in a museum XR project is: what happens to the visitor in the five minutes before and after? Work through it on a floor plan before anything is built.
- Where does it sit in the route? An experience that blocks a narrow corridor creates crowding. One at the end of a route, when people are tired, gets skipped.
- How long does it last? Shorter is almost always better. A focused few minutes that does one thing well beats a long experience that holds up a queue.
- How many people per hour? Work out throughput against expected footfall at peak times, not on a quiet weekday.
- Who staffs it? Front-of-house teams are already busy. If a station needs a dedicated person all day, that is a running cost to budget for, every year.
- What happens when it breaks? A dead headset or a failed tracking marker on a busy Saturday needs a simple recovery a non-technical staff member can do in a minute.
These answers usually shape the format choice more than the content does.
The collection is the content pipeline
The visual quality of a museum XR experience depends on the source material, and institutions often have more than they realise: photogrammetry or laser scans from conservation work, archive drawings, survey data, high-resolution photography and the expertise of their own curators.
Each source needs work before it runs on a headset or a phone. Scan data is far too heavy as captured and has to be simplified and retextured. Reconstructions need decisions about what is known and what is inferred, which leads to the most important content question in heritage work: how certain is this? Good reconstructions make the difference between evidence and interpretation visible, for example by showing surviving fabric differently from reconstructed sections, or by letting visitors see the sources. Curators should sign off every scene, and the schedule needs to allow for their review cycles.
Our guide to the XR 3D content pipeline explains how scans and models are turned into real-time assets, and what it means for budget when the source material needs heavy cleanup or has to be created from scratch.
Rights matter too. Check who owns the scans, the photography and any third-party models before they are used, and agree at contract stage who owns the optimized assets and the source project at the end. Software ownership and handover covers what to ask for.
Accessibility and language are part of the brief
A public institution has a broad audience, and XR can exclude people easily if access is an afterthought.
- Physical access. Seated options, experiences that work from a wheelchair height, no reliance on standing or walking around.
- Sensory access. Captions for all narration, audio description where visuals carry meaning, options that do not depend on colour alone.
- Alternatives. A non-headset version of the same content, such as a video or a screen-based view, for visitors who cannot or do not want to wear one.
- Language. Multilingual narration and text, ideally chosen once at the start rather than buried in a menu.
XR accessibility goes through these in detail, and it is cheaper to design them in from the first prototype than to retrofit them.
Running it for years, not weeks
Many museum installations are expected to run for the length of an exhibition or longer, on hardware handled by the public every day. Plan for that from the start.
- Kiosk mode. The device should boot straight into the experience and return to its start screen after each visitor, with no access to system menus.
- Device management. Headsets and tablets need updates, charging routines and a way to push new content without a technician visiting each one. Rolling out XR across many devices covers fleet management.
- Hygiene. Face covers, cleaning routines and spare units, built into the operating plan.
- Offline operation. Gallery Wi-Fi is often unreliable, so the experience should run without a connection and sync usage data when it can.
- Platform change. Headset models are discontinued and operating systems update on their own schedule. XR app maintenance explains how to build so those changes are manageable.
Measure what matters to the institution: how many visitors used it, where they dropped out, how long sessions lasted and whether queues formed. Keep any data collection minimal and anonymous, especially where children visit. XR data and privacy covers what headsets can capture and how to limit it.
What drives the cost
We will not quote a figure without a scope, because museum projects vary enormously. The main drivers are:
- The format: browser AR, a native app, staffed VR or shared mixed reality.
- How much 3D content is needed, and how much must be created rather than adapted from existing scans.
- The standard of historical accuracy and the number of curatorial review rounds.
- The number of languages and the depth of accessibility features.
- Whether it runs indoors on managed devices or outdoors on visitors' own phones.
- How long it must run, on how many devices, and who maintains it.
What an XR app costs explains each driver. A sensible first step is an XR proof of concept: one object or one scene, tested with real visitors in the real gallery, before committing to a full exhibition.
Where Null Studio fits
We are an AI-native software studio building across AI automation, custom software and XR. Two pieces of our XR work map closely onto heritage projects. Planes XR for ARCortex is location-based AR connected to OpenSky live flight data, running real plane simulations in the real world using AR and live maps, which is the same core problem as placing content accurately at an outdoor site. MR Camera is a multiplayer mixed reality environment where several people place and interact with 3D models in the same space, which is the mechanic behind shared group experiences. We also built ERIS XR for ARCortex, which blends digital and physical spaces in real time for firefighter pre-planning and training, and Nystag, clinical VR eye-tracking on the Vive Focus 3.
For multilingual narration and visitor questions, our AI work applies too: we build the AI voice and SMS systems behind Fortell, which handle calls and intake in more than 100 languages, and multilingual voice AI explains the approach.
If you are comparing studios, choosing an XR development company lists the questions we would want you to ask us as well.
A checklist before you commission
- What can the visitor not see today that XR would show them?
- Phone AR, staffed VR or a shared group experience, and why?
- Where in the visitor route does it sit, and what is the throughput at peak?
- Who runs it day to day, and what do they do when it fails?
- What source material exists, who owns it, and who signs off accuracy?
- Which languages and accessibility options are required at launch?
- How long must it run, on how many devices, and who maintains it?
- Who owns the source project and 3D assets when the work is done?
Planning an AR trail, a VR reconstruction or a shared XR experience for a museum or heritage site? Book a demo and we'll help you pick the right format and scope a first phase you can test with real visitors. See our XR work: Planes XR and ERIS XR for ARCortex, MR Camera shared mixed reality, and Nystag clinical VR.