SimStad Technical Guide

This deployment guide details the hardware and software architecture of the SimStad exhibition platform originally developed for Visualization Center C in Norrköping. SimStad bridges the gap between deep-learning spatial analysis and public outreach, allowing urban planners, municipal tech teams, and citizens to visualize complex city datasets over a physical landscape.

The core technology behind SimStad has since laid the groundwork for the Visual City Platform (2024–2026)—a nationwide initiative funded by FORMAS and Smart Built Environment. Utilizing deep neural networks for automated 3D reconstruction and semantic mapping, this ecosystem is designed to help Swedish municipalities independently spin up their own digital twins using advanced visualization and image analysis methods.

If you are a municipal IT specialist, exhibition technician, or developer looking to replicate or adapt this architecture for your own city, this documentation is for you. Learn more about the history and impact of the project on our main SimStad Project Page.

System Architecture

The SimStad exhibition is a modular system where map data, projection, and interactive screens communicate in real time over a localized web server.

3D Print

The physical 3D city model is the center of the exhibit. The model in Norrköping consists of 16x12 tiles printed in white PLA plastic. The same 3D city model is used virtually in the Unreal Engine application.

Read more about how we created the 3D model of Norrköping and how we printed it here:

Documentation.

Unreal Engine application

The Unreal Engine application is connected to the projectors and displays the map data onto the 3D print. It uses a copy of the 3D print model to simulate accyrate projections and realistic lighting.

User interface

A clean, web-based control application allowing visitors or presenters to toggle specific datasets (such as flooding vectors or noise pollution mapping).

Omni Communication Server

A custom backend infrastructure operating as a local database and WebSocket hub, handling all traffic between the control interface and the Unreal Engine render node.

Pixilab Blocks

Surrounding the exhibit are additional displays. These are used to display data legends and supplimentary videos to make the exhibit more immersive. These displays are synchronized using Pixilab Blocks. The interface sends signals to Blocks, deciding what content to be shown.

(Documentation)
Purchase Blocks

Live-data and plug-ins

Modular pipeline additions that ingest external real-time data APIs (e.g., live municipal traffic feeds) and stream them straight into Unreal Engine via Network Device Interface (NDI) layers. Explore Live Data Plugins.

WebSocket Protocol

To make the installation modular and future-proof, the communication framework between the control interface, the Omni server, and the Unreal Engine rendering pipeline relies on a clean, message-driven network protocol. If your municipality wishes to replace the frontend, substitute the rendering stack, or add custom input hardware, you can hook straight into our defined WebSocket Protocol Guide.