XR or Projection Dome. How to Choose a Simulator Display

Starting point

The question is not which technology is better

A projection dome and an XR headset solve the same problem: give the trainee an image of the environment they are supposed to act in. They do it in ways different enough that comparing them head to head, parameter by parameter, leads to poor decisions. A dome is infrastructure, a headset is station equipment. That distinction returns in every criterion below.

The right question is: what training objective must the system deliver, how many people have to pass through it per year, and what budget and floor space does the training centre have. The answers to these three questions usually settle the choice faster than any table of optical parameters.

How the two technologies work

Projection dome

A set of projectors displays the scene on a spherical or cylindrical screen surrounding a cabin or a station mock-up. The image is shared by everyone inside, requires nothing on the head and covers a wide field of view. The system needs a dedicated room of sufficient height, a multi-channel image generator, precise edge blending between projectors and ongoing maintenance: lamps or laser modules, convergence, geometry.

XR headset

Each trainee wears a display covering the full field of view, with head tracking. In the mixed reality variant, cameras pass through the image of real instruments, hands and cabin, while the outside world is synthetic. The trainee operates physical switches, sees their own hands, and has any terrain behind the cabin glass. Enterprise-class devices, for example Varjo XR-4, add eye tracking on top: a measurement of where the trainee is looking, used for assessment and for adapting the scenario.

Comparison across eight criteria

Projection dome and XR headset, decision criteria
Criterion	Projection dome	XR headset
Cost per station	high, driven by projectors, screen, image generator and the building	an order of magnitude lower, driven by the headset and a workstation-class computer
Floor space and infrastructure	dedicated room, height, cooling, blackout	desk-sized station or cabin, containerised setups possible
Time to deploy	construction design, installation, calibration	weeks from delivery to training
Mobility	none, the system is stationary	transportable between centres and units
Team training in one room	natural, everyone sees the same image, the instructor sees the trainees	networked stations see the same scene, people occlusion shows the real co-trainee inside the virtual environment, the instructor watches through a monitor view
Long sessions and comfort	nothing on the head, multi-hour sessions	headset weight and heat limit uninterrupted session length
Real instruments in view	yes, physical cabin in front of the screen	yes, in mixed reality through pass-through cameras
Measuring trainee attention and workload	needs external eye tracking, rarely used	eye tracking built in, data from every session

Costs are given qualitatively because they depend on configuration: number of projection channels, image generator class, number of XR stations and integration requirements. To run the numbers for your own case, use the calculator.

Where the dome remains the right choice

A fair comparison requires saying this plainly: there are use cases where the dome holds its ground and will keep holding it.

Multi-hour sessions without breaks. Having nothing on the head matters in continuous exercises, examinations and scenarios where equipment fatigue would distort the measurement of competence.
Populations sensitive to simulator sickness. The share of people who tolerate headsets poorly drops with every hardware generation, but it is not zero. The dome is neutral in this respect.
Existing infrastructure. A centre that already owns a dome with a current image generator rarely has a reason to abandon it. Complementing it with XR stations serving a different purpose usually makes more sense.

Where XR headsets have the advantage

Cost and number of stations. The budget of one dome buys many XR stations. If the bottleneck of a training centre is throughput, and in European armour and aviation programmes it usually is, the number of stations translates directly into crews trained per year. I describe this mechanism in the training hour gap analysis.
Floor space and mobility. An XR station fits in an office room, a container or a unit's hall. Training capacity can move to where the people are, instead of moving people to the building with the dome.
Mixed reality with real instruments. Pass-through lets the trainee keep their hands on the physical controls of the station, see the co-trainee next to them, and still operate in synthetic terrain. For operator stations, turrets and teleoperation of unmanned platforms this is now the natural configuration. People occlusion, introduced in Varjo Base 4.16, goes a step further: a real co-trainee is visible inside the virtual environment, so a crew training in the same room sees each other while looking at the same synthetic scene.
Measuring training outcomes. Built-in eye tracking yields data on the trainee's attention distribution and cognitive load from every session, with no extra equipment. What can be read from that data, I show in a research prototype.
Speed of scenario change. Changing terrain, time of day, weather or the whole platform type is a software change, not a rebuild of the room.

A separate problem: the motion platform

In the 2024 article I gave this issue the most space, and it remains valid. A headset tracks head movement, while a motion platform moves the trainee's whole body together with the station. A tracking system that cannot tell platform motion from head motion introduces visual artifacts and discomfort. The solution is compensation: tracking referenced to the platform's frame, or software that subtracts platform motion, both available today in mature integrations.

A practical note for specifications: if the simulator has a motion platform, the ability of the headset tracking system to compensate for platform motion should be an acceptance requirement, tested on motion profiles from the target scenarios, not a claim from a brochure.

The dome does not have this problem by design, because the screen stands still relative to the world. It pays for that with immobility and cost, so here too the choice comes down to the training objective, not to the superiority of one technique.

Conclusions

This is not an either-or choice Mature training centres build a layered structure: a dome or full-mission simulators where crew coordination and collective procedures are trained, and next to it a scalable layer of XR stations for individual, procedural and sustainment training, which consumes the most hours in any programme.
The criterion is training objective and throughput If the system's job is to push hundreds of people through a programme per year, the number of stations beats the parameters of any single station. If the job is certified team training in long sessions, the shared screen has its justification.
Measuring outcomes is no longer an add-on The requirement that a training system delivers data on trainee progress appears in a growing number of procurements. Eye tracking built into the headset produces that data almost as a by-product. In a dome it has to be collected with separate instrumentation.
Precede the decision with arithmetic, not a demo A demo shows image quality, but it does not show the cost of a training hour or annual throughput. Those two numbers, computed for your own data, settle most technology disputes faster than any showcase.

Run your numbers Discuss your project

Sources and caveats

Where this comparison comes from

Dawid Dobies, VR/XR vs. Projection Dome Systems: Which display technology will drive the future of simulation?, LinkedIn, September 2024. The first version of this comparison, written from a motion platform integrator's perspective.
The author's project experience: motion platforms and XR integrations in vehicle simulators, and XR deployments in defense training across Central and Eastern Europe and the Nordics.
Publicly available specifications of projection systems and enterprise-class XR headsets.
Varjo, Release notes Varjo Base 4.16, July 2026: People occlusion mode (Beta) showing real human figures inside the virtual environment.

This analysis is illustrative and does not replace the assessment of a specific procurement. Costs are given qualitatively because they depend on configuration. The author works professionally in the XR market, which should be said plainly; the criteria in this text can be applied against headsets just as well as in their favour, and that is their purpose.