Why Replicity
Cities aren't scanned into Unreal. They're modelled into it.
The Process
Imagery in. Procedural cities out.
Replicity specializes in creating concrete spatial data from arbitrary satellite and street-level imagery. We don't depend on exclusive LiDAR fleets or perfect capture conditions — we turn the imagery you can actually get into geometry you can ship.
From that data, we procedurally model large-scale real-world locations inside Unreal Engine: streaming-friendly LODs, editable structure, and performance budgets that hold at city scale — not one-off showcase meshes. Because the environments are built for Unreal, you can visually customize materials, lighting, and look — and keep editing the city after handoff without fighting a frozen scan.
Delivery is OpenUSD-first for interchange, with Unreal Engine as the primary runtime. The result is a city you can light, stream, restyle, simulate, and iterate — not a brittle photogrammetry dump.
The Data
What you get from Replicity.
- Imagery → structured data
- Satellite and street-level sources converted into concrete spatial geometry
- Procedural city systems
- Large-scale locations modelled procedurally for editability and scale
- Unreal Engine delivery
- Native UE5 packages tuned for streaming, LODs, and real-time performance — visually customizable and easy to edit in-editor
- OpenUSD interchange
- Primary handoff format for pipelines that span DCC tools and stages
- Georeferenced worlds
- EPSG:4326 default with custom CRS options for simulation and twins
The Guarantee
What we optimize for.
City-scale
Procedural coverage built for districts and full metros
UE5-first
Visually customizable Unreal environments that stay easy to edit after delivery
Any imagery
Expertise turning arbitrary satellite and street sources into usable data
Specs
How we compare.
| Replicity | Photogrammetry dumps | Manual city modelling | |
|---|---|---|---|
| Input | Satellite + street imagery | Dedicated capture flight | Reference plates / maps |
| Approach | Procedural modelling | Dense mesh reconstruction | Hand modelling |
| Scale | District → metro | Site → neighborhood | Block → precinct |
| Primary DCC | Unreal Engine · OpenUSD | Mesh tools · FBX/OBJ | DCC of choice |
| Runtime fit | Streaming-ready UE5 | Often needs retopo | Depends on artist |
| Edit & look-dev | Easy in-editor visual customization | Hard to restyle cleanly | Artist-dependent |
Case Studies
Results in production.
Chicago Loop
Unreal game studio
12× faster
city block turnaround vs. hand modelling
Singapore CBD
Simulation platform
OpenUSD + UE5
single source of truth across tools
Dubai Marina
Virtual production stage
1 imagery set
procedural city rebuilt for LED volume use