Anchor Critical Information to the Real World

Don't just look at your equipment - see inside it. Our 3D Object Tracking locks digital content to physical objects with sub-millimeter precision, allowing you to visualize live IoT data, maintenance history, and assembly guides exactly where they are needed. MultiSet turns your 3D model (GLB/CAD/scan) into a production-ready “object  anchor” in minutes—no markers, no fiducials. Publish once, then deploy  to Unity, WebXR, iOS and Android with enterprise-grade privacy options  (public cloud, private cloud, or fully on-prem). Create object anchors  in WebXR with ModelSet.

Smiling man wearing glasses and a gray sweater against a light background.
Request For Demo
Start Free

What is Object Tracking?

Object Tracking recognizes complex physical equipment based on its 3D shape. This allows you to point a camera at a machine and instantly see digital repair guides, safety warnings, or live data "stuck" precisely to the specific bolts, levers, or screens you need to interact with—no QR codes required.

How it works

UPLOAD 3D FILE
GLB or GLTF
Author at true scale (1 unit = 1 m)
Align up = +Y
PICK TRACKING TYPE
360° View
Side View
UPLOAD, PROCESS, PUBLISH
Enable localization and tracking in mobile, tablets, headsets and robots
Why teams choose Object Tracking
Futuristic industrial scene with holographic robotic arm and augmented reality tracking system icons connected to various platforms and cloud security options.
Sub-5 cm accuracy. Rock-solid anchoring with low drift in busy, changing environments.
Scan-agnostic. Works with CAD, GLB/GLTF, or meshes from your reality-capture pipeline.
Fast to live. Create an Object, upload, process, and ship—typically in under 10 minutes.
Built to scale. Indoors or outdoors, multi-floor, any lighting.
Deploy anywhere. Unity, WebXR (browser), iOS, Android, Vision Pro, Quest.
Enterprise security. Choose cloud, private cloud, or on-prem/self-hosted.
Cross-platform SDKs. Unity • Native iOS & Android • WebXR • ROS bridges.
Scan-Agnostic Object Capture
Scan objects with Polycam, Scaniverse or any scanner that outputs GLB &  PLY with textured meshes to MultiSet; Vision Fusion normalizes them into object trackers on one high-fidelity map on a unified coordinate system so teams can use the hardware they already own.
Bring your Object
Diagram showing PolyCam in the center connected to six applications: handheld scanner, photogrammetry object scanning, assembly guidance, AR maintenance, interactive packaging, and tangible gaming.

Track Any Object In Camera FOV

Build real, shippable AR that recognizes the exact object and locks content precisely where it belongs - no markers required. With sub-5 cm anchors from our markerless object tracking SDK, experiences stay rock-solid across Unity, WebXR object anchors, iOS, and Android, so one project ships everywhere.
Maintenance & Inspection
Recognize machines and pin SOPs, torque specs, and checklists right where work happens.
Person holding a digital tablet displaying augmented reality interface for monitoring industrial pipes in a facility.
Learn More
Warehouse & Training
Identify tools/parts, guide picks, and standardize onboarding with in-place AR cues.
Warehouse worker in a yellow hard hat scanning boxes on shelves using a handheld scanner and holding a tablet, with digital interface overlays visible.
Book Pilot
Field Service
Heads-up diagnostics and step-by-step repairs—works offline in on-prem mode.
Industrial pipe with pressure gauge and augmented reality overlay displaying 'MAINTENANCE NEEDED'.
Test Now
Retail, Museums & Exhibits
Interactive product and artifact experiences—no QR codes or markers needed.
Collection of ancient ceramic pots with one labeled A describing a ritual storage pot from the late 3rd century BCE used in seasonal harvest rituals.
Learn More

Technical Specs

Input formats: GLB/GLTF (CAD and scans convertible)
Tracking modes: 360° View, Side View
Precision: Sub-5 cm (typical in recommended conditions)
Environments: indoor/outdoor, dynamic lighting, multi-floor, large-scale
SDKs: Unity, iOS, Android, WebXR; ROS adapters available
Privacy: cloud / private cloud / on-prem; opt-ins

Create Object Anchors in WebXR With Object Tracking

Try for free
Frequently asked questions
Basics
Setup
What does “markerless object tracking” mean?

No stickers, QR codes, or fiducials. The system uses the object’s geometry and texture to recognize and track it in 3D.

What’s an “object anchor”?

In WebXR, an anchor is a pose (position + orientation) the system keeps aligned to the real world as tracking updates. Object anchors are anchors tied to a recognized object rather than an arbitrary spot.

How does ModelSet's object tracking function?

Upload a textured 3D model to the MultiSet cloud, which then generates an optimized tracking map. The device SDK recognizes the object and maintains its pose locally, ensuring responsive and low-latency AR.

Do I need textures if my source is CAD?

Yes. CAD is often untextured; convert to a polygonal mesh and apply a realistic texture so the tracker has visual features to lock onto.

Which objects track best?

Asymmetrical objects with rich features (logos, panels, labels, varied geometry) perform best. Highly symmetrical or feature-poor/reflective items are challenging.

What’s the relationship between ModelSet object anchors and WebXR “anchors”?

ModelSet provides the stable pose tied to a detected object; you then maintain alignment using WebXR Anchors as the session evolves. Think: ModelSet → initial precise pose, WebXR Anchors → keep it glued as tracking updates.

How do I convert CAD to GLB/GLTF for object tracking?

Export a polygon mesh from your CAD tool (or via a DCC like Blender), apply a PBR base color/diffuse texture, verify real‐world scale (1 unit = 1 m), set +Y up, and export to .glb/.gltf. Keep the file lean (ideally <20 MB), with clean UVs and a single material when possible.

Which file formats are supported for object tracking?

Export .glb or .gltf. (CAD and scans should be converted to polygonal GLB/GLTF before upload.)

What are the limitations on object and file sizes for tracking?

Ideal object sizes are 1 ft to 25 ft; keep the model file under 20 MB for smooth processing. Typical range is 0.3 m to 7.5 m on the longest dimension. Smaller objects may lack features; very large objects work best when the model captures distinctive local details.

Can Object Tracking track moving or deformable objects?

Today Object Tracking is designed for static, rigid objects during the AR session. The user/camera can move around it but lossless tracking with the same level of robustness as MultiSet VPS.

What’s the difference between 360° View and Side View?

360° View (default): recognize from any angle (top/bottom/sides). Side View: optimized when the object is viewed mostly from the sides; usually processes a bit faster.

How long does processing take?

Typically under 10 minutes, depending on model complexity. You’ll see status change to Active once ready.

Do I need to ship the 3D mesh in the final app?

In Unity, the reference mesh helps authoring and content placement; remove it from the final build (it’s not used during runtime localization).

Request a Demo

→

Start Building

→
Pages
HomePricingBlogGet In TouchPivacy Policy
Platform
VPSObject Tracking3D Mapping
Developers
TechnologyDocumentationVuforia AlternativeASA AlternativeStatusCommunity
contact us
28 Geary Street
STE 650 Suite #371,
San Francisco,
California 94108, USA
Follow us
© 2026 Multiset AI
All rights reserved
Terms of use
·
Privacy Policy