What's the difference between a digital twin and a VPS?

A digital twin shows you a space. A visual positioning system (VPS) tells a device where it is inside that space, in real time. MultiSet is the VPS layer that makes a twin operational for cameras, robots, and agents.

What reality-capture formats does MultiSet support?

E57 from Matterport, Leica, NavVis, XGRIDS and FARO; Matterport MatterPak; textured meshes; raw LiDAR; 360 video; metric-scaled Gaussian splats; and 3D tiles, including public and third-party sources.

What makes MultiSet scan-agnostic?

We accept the widest range of formats and normalize them into one coordinate system - so you're never locked to a single scanner or vendor. New map versions and new MapSet sections can use different hardware and formats.

Can I localize a camera or robot against 3D data I already own - point cloud, mesh, BIM, digital twin, splat, or GIS?

Yes - in real time, with no re-scan.

Do I have to capture anything myself?

No. Use a scanner you own, capture with a phone, have our partners capture for you, or localize against public and third-party 3D data that already exists - 3D tiles and open geospatial sources.

How do I make my reality capture data actionable?

Feed the scans you already own into MultiSet. It turns them into a live map devices localize and track against - so the capture drives action, not just viewing.

Can I export maps to other spatial tools?

Yes. Maps and MapSets export as GLB, USDZ or OBJ meshes, PLY or SPZ splats and a PCD point cloud, letting you reuse geometry in BIM, game engines or digital-twin analytics platforms - no vendor lock-in.

How does MapSet enable large‐scale coverage?

MapSet stitches many smaller scans into one seamless “mega-map.” It preserves the detail of each fragment while providing a single coordinate system, so users experience uninterrupted navigation across rooms, floors and outdoor areas.

Scan-Agnostic 3D Mapping - Any Capture to VPS

Agnostic on every axis.

Scan-agnostic: what you feed in

Any format or scanner builds the map: E57, LiDAR, 360 video, textured meshes, MatterPak, Gaussian splats. Keep your hardware.

Sensor-agnostic: what localizes

Any device localizes and tracks: phone, headset, robot camera, LiDAR, 360, drone today; anything that sees tomorrow.

Map-agnostic: what you localize against

Any 3D data is a map: point clouds, meshes, 3D tiles, splats, GIS layers, digital twins. Even data you didn't capture.

Vendor-agnostic: who you're tied to

Any capture vendor, any cloud, any deployment. Your data never becomes our model. No lock-in.

Three ways to get a VPS map

Bring Any Scan

Use the scanner you already own. Keep your hardware. Keep your workflow.

Supported Formats

Managed Capture

No scan yet? We'll capture it for you through our global network of capture partners.

Talk To Us

Capture It Yourself

Prototype in minutes. Scan on the spot with the MultiSet app on iOS Pro.

Get iOS App

Different scanner for every zone. One map underneath.

No 3D data yet? Capture it.

Use a drone outside, an AMR through corridors, a survey rig where detail matters, a phone for the rest. Every scan merges into a single VPS-ready map on one coordinate system. Choosing the right tool per zone is the only real decision.

Learn about reality capture

Three levels of an industrial facility drawn as cutaways held in one shared coordinate frame, a technician with a tablet on the middle level, and a red route running between the levels

Point clouds, E57, textured meshes, 360 videos, PLY, 3D Tiles and GLB or glTF ingested into MultiSet, which exports a machine-readable VPS with SDKs and APIs alongside a human-readable visualization as GLB, PLY, SPZ and 360 tours

Already have 3D data? Localize against it.

Point a camera at a space you've captured before and MultiSet places the device inside it, in real time.

Point clouds, meshes, tiles, splats, GIS layers you already own. No re-scan, no conversion project. The data you've been collecting for years becomes a live positioning layer.

Learn about localization

One capture. Two ways to use it.

Every MultiSet map is readable two ways at once: a photoreal view your team can see and review, and a machine-readable map your devices, robots, and agents localize against in real time. Capture once. Serve people and machines from the same source of truth.

Your scan. Your cloud. Your rules.

MultiSet runs the same everywhere. Your capture vendor stays your choice. Your data never becomes someone else's model. That's what makes MultiSet a neutral layer instead of another walled garden.

Dedicated cloudPrivate VPCOn-DeviceAir-gapped

MultiSet iOS Mapper

Turn any LiDAR-enabled iPhone or iPad into a complete VPS prototyping kit. Scan a space to build a VPS map in minutes, author and share AR experiences instantly with Content Spaces, capture simulation data, and validate localization and object tracking success on the spot. The fastest way to test MultiSet before you build.

Start Free

Four smartphones displaying a 3D industrial facility model with interconnected pipes highlighted in purple and dark-themed app interfaces with tools and maps.

Composite image showing digital wireframe overlays on photos of an industrial facility, shopping mall interior, suburban neighborhood, and urban cityscape.

No size limit. If a camera can see it, MultiSet can map it.

MapSet stitches every scan across floors, buildings, indoors and out into one coordinate system, so a map grows as large as you keep capturing and navigation never breaks at a doorway. Change one area without remapping the rest.

Combine & merge multiple scanners and file formats to build limitless maps.

MapSets

Supported scanners and file formats

MultiSet ingests reality capture from every major scanner and format:

Provider	Devices	Capture type	Formats ingested
Apple	iPhone Pro, iPad Pro (MultiSet app)	LiDAR	Native LiDAR
Matterport	Pro2, Pro3	Laser + photo	E57, MatterPak
Leica Geosystems	RTC360, BLK360, BLK2GO	Terrestrial & mobile LiDAR	E57
NavVis	VLX 2, VLX 3, MLX	Mobile mapping (SLAM)	E57
FARO	Focus, Focus Premium	Terrestrial LiDAR	E57
XGRIDS	PortalCam, Lixel K1, K2, L2 Pro	Mobile LiDAR + Gaussian splat	E57, Gaussian splat
Insta360	X4, X5	360° video	360 video
Antigravity	A1 (360° drone)	Aerial 360° video	360 video
Polycam	iOS / Android app	Phone scan	GLB, PLY (textured mesh)
Any E57-compatible scanner	-	Terrestrial / mobile LiDAR	E57
Any metric-scaled 3DGS tool	-	Gaussian splat	Gaussian splat
Public / third-party 3D	-	Existing 3D map data	3D Tiles

Per-scanner guides: Matterport, NavVis, Leica, FARO and XGRIDS. For aerial capture, see drone to VPS.

Inputs and outputs

How to convert reality capture data into a machine readable localization map

One capture produces two different things. A machine-readable map, the feature representation a camera localizes against, and human-readable artifacts, the mesh, splat, panorama tour and point cloud people look at. Most of the market produces one or the other.

On how the human-readable artifact is produced and where it differs from a survey deliverable, see Gaussian splatting versus photogrammetry. Terms used here are defined in the glossary.

For what a processed map hands back, in which formats and in which coordinate frame, see map outputs.

Input	Source hardware or tool	Produces	Notes
E57	Matterport, Leica, NavVis, Faro, XGRIDS	VPS map, point cloud	The industry interchange format. Most existing scans are already here
MatterPak	Matterport	VPS map, mesh	The bundle Matterport exports for downstream use
GLB / PLY	Polycam, Scaniverse, CAD	VPS map, mesh	Textured meshes
Gaussian splat (.ply / .spz)	Any 3DGS pipeline	VPS map, splat	Must be metric-scaled. A scale-free splat is a picture, not a map
360 video	Insta360 and similar, raw or equirectangular MP4	VPS map, panorama tour	One walkthrough, or one flight
Native LiDAR	iPhone and iPad Pro, via MultiSet Mapper	VPS map, mesh	No professional hardware required

One map. Every device. Any capture.