Home / Case studies / A national museum in Central Asia · Digital exhibition
Multilingual digital exhibitions with 3D content that runs smoothly on ordinary devices.
A national museum in Central Asia that needed its collection and historical narrative digitised and presented to a multilingual audience.
There were three difficulties: language, which required supporting the local language, Russian and English at the same time; hardware, since venue terminals are modestly specified and cannot rely on high-end GPUs; and network, since connectivity is unstable in some areas and content has to run locally.
Delivered a digital exhibition system with 3D exhibit display, multilingual guides, touch interaction and large-screen sync.
3D content is made lightweight: after model compression it runs smoothly on ordinary devices without a discrete GPU.
Content per language is managed independently; switching language does not change page structure.
Content is separated from code, so the venue can replace exhibit material itself later instead of calling developers every time.
The table below lists this project's technical elements so clients can verify them independently.
Project stories without constraints are fiction. This section sets out four constraints: time, site, staffing and compliance.
These constraints directly shape the solution; every problem under 'Key challenges' below is forced out by one of them.
It is not “we worked hard”; it is “what the problem looked like, how we solved it, and what it cost”.
This section is for peers. If you are doing something similar, these points will save you a few detours.
Three tricks for running 3D on low-end devices are far cheaper than new hardware.
Every multilingual project should work this way to avoid N times the maintenance.
Any unattended exhibit should have one.
Talking only about successes is unconvincing. This section covers what we got wrong or did too late.
Doing it again, we would specify equipment models and performance in the contract instead of a vague mention of several all-in-one machines. Differences in performance forced a full rework of the design.
We underestimated how unstable the local power supply was. We later learned there are two or three outages a week; had we known, the first version would have included a UPS and a watchdog.
Manual proofreading of the trilingual content was not sufficient. Machine translation drifts on technical terms, and several inaccurate phrasings were reported from site, so manual proofreading was added afterwards.