Baaz turns aerial survey and open geospatial data into measured, queryable city models - geometry you can trust, corridors you can plan, and live environmental context, all streaming to a browser.
A digital twin is only useful if you know which parts of it are real. Ours labels every value as measured, live, official or simulated - and where an independent reference exists to check against, we publish our error rather than claiming accuracy we have not tested.
What we can build
This is the heart of the twin - finish it and you know exactly what you would receive. It is built to answer plain questions directly: how tall is that building, what has been built since 2016, can we fly this route, which structures flood at this level, does this rooftop get enough sun, and how dense is this district.
The Base Model
A twin starts as geometry. We reconstruct the built environment into solid, individually addressable objects rather than a decorative mesh - buildings with footprint, height, volume and estimated storeys, bare-earth and surface terrain, classified point clouds that separate ground, vegetation, structures and wires with a confidence value on every point, and street assets like poles, trees and signs extracted as discrete, countable objects. Every object carries its attributes, so the twin behaves like a database you can query, not a picture you can only orbit.
The City Around It
Geometry alone does not read as a city. We layer in real surveyed context - road, rail and metro alignments, water bodies and green space, land-use and zoning per parcel, hospitals, schools and civic infrastructure, and neighbourhood boundaries with per-district statistics - so the model reflects the place, not just its buildings.
Time Axis
Where a multi-epoch archive exists, the twin gains a time axis. We animate growth between two dates, measuring what was actually built rather than modelling what might have been. Where an archive is too inconsistent to support the claim, we leave the feature out rather than ship a number that will not survive scrutiny.
Analysis on the Measured Surface
Direct answers from the height surface, no extra survey required - drone corridor and BVLOS planning with recommended cruise altitudes and binding obstructions, line-of-sight cuts for radio and sensor placement, annual sun and shadow exposure, and flood-exposure modelling at specific water levels. Each runs on the same measured surface rather than an assumed clearance.
Live & Official Data
Live weather and air-quality feeds, statutory property rates and official identifiers from government sources, and IoT sensor or live-video integration - each badged for what it is, so a live reading is never confused with a modelled one.
The Operations Layer
A professional operations console for control-room use - district statistics, historical charts, alerting thresholds, and a searchable index across every feature. Where a client has real telemetry it connects here. Where they do not, we demonstrate the interface with a clearly badged simulation, never presented as a feed we do not have.
Delivery & Standards
Browser-native delivery with no plugins or proprietary viewers, a REST API queryable by geometry, height or class, and open standards throughout - 3D Tiles, LAZ and CityJSON. Deploy hosted, on-premise, or fully offline for secure sites.
Technology stack & geospatial pipeline
The survey sources, modelling standards, and delivery layer we use to turn raw aerial and open data into a measured twin that streams to a browser.
Survey & Terrain
Sources and reconstruction for measured ground, surface and canopy models.
Airborne LiDAR
Photogrammetry
Satellite-derived height models
Surface reconstruction
Geospatial & Modelling
Open formats and modelling standards that keep the twin interoperable.
OGC 3D Tiles
CityJSON
glTF
OpenStreetMap
Delivery & Ops
Streaming, APIs and reproducible pipelines that put the twin in the browser.
WebGL Streaming
REST APIs
Reproducible pipelines
Environmental services
Who it's for
The twin is built for teams that need measured answers about a place, not a decorative model of it.
Drone & UAV operations
Corridor design, obstacle clearance, ground-station siting, and BVLOS planning support - all read off the measured height surface.
Urban planning & government
Massing studies, zoning analysis, growth monitoring, and development control on a model that reflects what is actually there.
Infrastructure & utilities
Asset inventory, clearance checks, and network or sensor placement against classified, countable objects.
Real estate & investment
Built-form analysis, density and volume studies, and official valuation context in one queryable model.
Disaster resilience
Flood-exposure modelling, emergency-service coverage, and scenario planning on measured terrain.
How we work
Scope & Source
We establish what data actually exists for your area - flown LiDAR, satellite height models, municipal records, or nothing at all - and tell you what each option can and cannot support before any work begins.
Build
The pipeline ingests, reprojects, classifies and reconstructs. Every stage is reproducible, so the twin can be rebuilt from source at any time rather than existing as a one-off export.
Validate
Where an independent survey exists to check against, we measure our error and publish the figure per building. Where none exists, we say so plainly instead of quoting a number we cannot stand behind.
Deliver & Extend
The twin ships as an application and an API. New analysis runs on the same measured surface, so the next question rarely needs new data.
Why Baaz for a geospatial digital twin?
We build twins you can trust the numbers in - every value labelled for what it is, and our error published where we have measured it.
We publish our error
Most twins cannot tell you how much to trust them. Ours can, per building - every structure coloured by its measured deviation from an independent survey. That figure goes on the screen, not in a footnote.
Real and modelled never blur
Every value carries a label - measured, live, official or simulated. Geometry is measured, weather and air quality are live, property rates are government figures, and anything modelled says so. A screenshot cannot be mistaken for telemetry we do not have.
Built for places without LiDAR
Large parts of the world have no open aerial survey. We built the same pipeline to run on open satellite and footprint data and proved it at metropolitan scale. When a flight is justified, the higher-fidelity path is already there - same tooling, more detail.
Drone-native by design
We are an autonomy company first. Corridor analysis, clearance and line of sight are not add-ons here - they are the reason the height surface is built to be accurate in the first place.
Where Baaz applies this stack
These engagement lines routinely pair with the toolchain above-from first release through optimisation and runway.
Geospatial Digital Twin - Frequently Asked Questions
No. We run two paths: aerial LiDAR where it exists, and satellite-derived heights with municipal or ML-derived footprints where it does not. The LiDAR path gives roof geometry, street assets and a stated accuracy figure. The open-data path gives you a full city quickly and at a fraction of the cost. We will tell you which fits before you commit.
Where an independent survey exists to check against, we measure our error and publish it per building. Where no independent reference exists, no honest accuracy figure can be quoted, and we will say that rather than invent one. Satellite-derived heights carry a known bias on slender towers, which we state on screen.
We have built twins from a single district to a full metropolitan area. The area of interest is a tile grid rather than a fixed extent, so scale is a question of compute budget, not a rewrite.
Yes - that is a core use. Corridor analysis returns cruise altitude above ground and above sea level, the tallest obstruction and where it sits along the route, the terrain profile, and sensitive sites near the path, all from the measured surface rather than an assumed clearance. It is planning input and does not replace airspace authorisation.
Weather and air quality are real feeds. Geometry, terrain, land use and proximity are measured. Property rates come from government sources. Traffic, utility load and camera imagery are simulated for demonstration unless you have real feeds to connect - and they are badged as such throughout.
Everything is standards-based, with a REST API over the model and open formats for the geometry. There is no proprietary format and nothing locked to our hosting. The twin can run entirely offline for secure sites.