Robotics / Spatial AI
A scalable indoor training corpus for perception, navigation, and spatial reasoning.
Roomza licenses structured hotel-room data, visual assets, and spatial captures across rooms at more than 6,000 hotels for robotics and spatial AI teams building models that need to understand real furnished interiors.
Why hotel rooms
Hotel rooms are one of the most repeatable indoor environments in the world. They contain beds, desks, bathrooms, windows, lighting, doors, storage, soft goods, reflective surfaces, tight paths, and accessibility variations across a large variety of real-world layouts. That makes them useful for training and evaluating models that need to understand furnished interior space.
Available modalities
| Modality | Examples |
|---|---|
| Structured room data | Dimensions, layout, room attributes, condition, amenities |
| Visual data | RGB stills, walkthrough video, object-level room imagery |
| Spatial data | Floor plans, spatial captures, room geometry, fixed-object placement |
| Context metadata | Market, property type, room category, renovation and condition signals |
| Accessibility metadata | Door clearance, bathroom layout, mobility-relevant features |
Spatial captures may include floor plan signals, room geometry, fixed-object placement, or other capture outputs depending on the property and collection method.
Use cases
| Use case | What Roomza enables |
|---|---|
| Indoor perception | Train models on real furnished rooms across varied layouts |
| Navigation | Understand paths, obstacles, furniture placement, and room constraints |
| Accessibility robotics | Identify mobility-relevant room features and barriers |
| Synthetic data validation | Compare generated rooms against real hotel-room distributions |
| Interior-state models | Track condition, maintenance signals, and guest-facing quality |
What you get
Repeatable interior category
Hotel rooms are a scoped, diverse, real-world interior category sampled across price points, layouts, brands, and geographies.
Visual and spatial assets
RGB stills, walkthrough video, and spatial captures available under the Full license where capture rights and property permissions allow.
Room-level metadata
Dimensions, layout, fixed objects, finishes, lighting, accessibility flags, condition signals, and amenity data.
Simple corpus pricing
Per-room pricing that scales with corpus size, not per-seat or per-query usage.
Pricing
Core: structured room data only.
Full: Core plus licensed images, video, and spatial captures where available.
| Rooms licensed | Core (structured only) | Full (images, video, spatial) |
|---|---|---|
| Under 100K rooms | $0.75 / room / yr | $5.00 / room / yr |
| 100K–500K rooms | $0.60 / room / yr | $3.50 / room / yr |
| 500K+ all-access | $0.45 / room / yr | $2.00 / room / yr |
Reference package: Full corpus across the Roomza network with structured metadata, visual assets, and spatial captures scoped to approved training use. At the all-access tier of $2.00 per room per year, full-corpus reference licenses run into seven figures depending on scope, exclusivity, and refresh requirements.
Licensing and provenance
Spatial and visual assets are licensed under capture-method-specific terms, with provenance metadata available during diligence.
- Corpus
- Room-level data across more than 6,000 hotels
- Provenance
- Record-level source, date, and rights basis
- Licensing
- Commercial license tailored to use case
- Delivery
- Bulk files or REST API
- Refresh
- Annual, quarterly, monthly, or real-time by license