Skip to content

Deployment Options

realvirtual is delivered as a static set of browser files. You can host that set on your own web server, a CDN or infrastructure inside the machine network. You do not need a realvirtual cloud service to render and operate a prepared standalone twin.

Choose the deployment model from the machine owner’s availability, privacy and network requirements. Live PLC data and optional local services can be added through realvirtual CONNECT, while the viewer and model remain under your hosting control.

There is no application server, no database and no runtime service anywhere in this list. The viewer is a folder of static files, and everything below is a different way of putting that folder in front of a browser.

What runsWhat you maintainLive PLC data
Hosted on the webA web server or CDN serves the filesA hosting contract and a URLThrough a gateway at the machine
Local web serverYour own server, an IPC, an intranet shareThe server, inside your networkThrough a gateway in the same network
One-click executableThe gateway serves the viewer itselfOne file on one PCBuilt in
Local developmentThe Vite dev server on a developer machineA Node.js toolchainThrough a local gateway

Upload the built folder to any HTTPS host: a CDN, a hosting provider, a company web server or an S3 bucket. The browser downloads the viewer and the GLB, and everything else happens on the visitor’s device.

Users can open one link on any device without installing anything. For confidential models, consider that anyone who can reach the URL can download the model.

Serve the same folder inside your network from an intranet server, machine IPC, plant server or NAS. Nothing leaves the site, and no external hosting provider is involved. An air-gapped network works the same way because the files need no internet connection.

This is the usual choice for plant-wide monitoring and for anything that must keep working when the internet does not.

realvirtual CONNECT carries the viewer inside its executable and serves it on http://localhost:5100. Download the file, run it and open the browser. You need no web server installation, Node.js or build step. The same process includes the live PLC connection.

This is the quickest way to run a 3D HMI on a machine PC. It also suits customer delivery: one file that runs offline and stays unchanged for as long as it is archived. Details: where the viewer comes from.

For building and adapting a twin, the viewer runs from source on a developer machine: clone the AGPL repository, npm install, npm run dev, and the Vite dev server serves the viewer at localhost with hot reload. Models dropped into the models folder appear in the selector.

The same source produces the folder you later upload or embed, so the development version matches what you ship.

  • Frozen delivery — archive the tested viewer, model and documentation together as one versioned, self-contained release. See below.
  • Open source — the viewer is AGPL-3.0; self-hosting is explicitly allowed, and what the licence asks in return is described under licensing the viewer.

Regulation (EU) 2023/1230 applies from 20 January 2027. For digital instructions, it requires a way to download, print and store the documentation offline, and it places long-term availability responsibilities on the machine manufacturer. Online documentation must remain available for at least ten years after the machine is placed on the market, or for the expected machine lifetime when that is longer.

A frozen release supports this responsibility because the complete tested delivery can be archived, reproduced and moved to replacement hosting without depending on a continuously changing application service. It does not by itself establish regulatory compliance: the machine manufacturer remains responsible for the access information, retained version, availability period and any required paper or safety information.

  1. Decide whether the twin must be public, intranet-only or air-gapped.
  2. Build and test one complete viewer, model and documentation file set.
  3. Publish that unchanged set to the chosen static host and record its release version.
  4. Verify the final URL, browser operation, downloads and offline-storage path from the target network.
  5. Archive the exact delivered files and define who maintains their long-term availability.