What we are building next
We believe in transparency about product direction. Here is what our engineering and research teams are working on — and what is coming over the horizon.
2025 Roadmap
Active development
Spatial Foundation Model
In DevelopmentA large-scale AI model pre-trained on 500M+ labelled geospatial features across cadastre, land cover, infrastructure, and urban morphology datasets. Enables fine-tuning for any spatial classification task with 10× less labelled data than training from scratch.
Real-time Cadastral Change Detection
BetaAutomated alerts when cadastral boundaries, land use classifications, or parcel attributes change — triggered by satellite change detection or direct registry feed. Land administrators receive a notification and review queue within hours of any change.
Semantic 3D City Model API
In DevelopmentQuery 3D city objects by semantic class via a REST API — "give me all buildings over 20 storeys within this polygon" or "show me all roads within 100m of this flood zone." CityGML-aligned schema with OGC API Features endpoint.
Federated Spatial Data Exchange Protocol
PlanningA protocol allowing government agencies to share spatial datasets across organisational boundaries without centralising data — queries execute at source, results federated at query time. Cryptographic provenance tracks every data access.
Natural Language GIS Query Interface
PlanningAsk your spatial data questions in plain English. "How many residential parcels are within 500 metres of a flood-prone area?" generates, executes, and visualises the PostGIS query automatically — no SQL required.
2026 Vision
The longer horizon
These are the big bets — ambitious capabilities that will define the next generation of the Airfree platform.
Autonomous Survey Processing
Full drone-to-parcel-boundary workflow with no human intervention. Drone flies, data uploads, AI processes, cadastral boundary extracted, quality-checked, and submitted to the registry — all automatically. Target: 90% reduction in field-to-registration time.
Planetary-Scale Digital Twin
A continuously updated 3D model of the entire Earth, refreshed from satellite imagery, IoT sensors, and crowd-sourced data. Query any point on Earth for its current state — land cover, building footprint, infrastructure, and environmental conditions — in real time.
Sovereign AI Inference
Run Airfree spatial AI models fully on-premise with zero cloud dependency. Models packaged as container images with offline licence validation — air-gapped classified environments can access the full Airfree AI capability stack without any internet connectivity.
How we prioritise
Our three guiding principles
Customer co-design
Every item on the roadmap started as a conversation with a customer. We run quarterly co-design sessions with government and enterprise customers to validate priorities before committing engineering capacity.
Government partnership
Major capability investments are co-funded or co-developed with government agencies. This ensures the platform solves real operational problems — not hypothetical ones — and accelerates deployment through shared risk.
Open standards first
Every new capability is built on or contributes to an open standard — OGC, STAC, GeoParquet, OGC API. We never build a proprietary format when an open one exists or can be created.
Join our beta program
Get early access to capabilities in development, provide direct feedback to the product team, and shape the platform before general release.
Apply for beta access