Oct 22th 10:00 AM ET
What Happens When AI Understands Reality?
Register Now

Visual Asset Management with semantic layer

RealityAssets™ in RealityPlatform™ lets you tag equipment directly in your reality capture, fill it with your company's own asset templates, link it to the system that already owns the data, and keep every tag when the site gets rescanned.
Visual asset management, explained

From a scan full of shapes to a 3D asset library you can query

A scan knows geometry. RealityAssets™ technology knows it’s pump P-1043.

  • Tag equipment right inside the scan: draw a box, click with the Magic Wand, or trace a pipe run. A single valve or a whole conveyor takes a few clicks.
  • Give every asset its real identity: company asset types bring the same fields to every site, filled by hand or pulled live from SAP, Maximo or AVEVA PI.
  • Keep every tag when the site changes: assets sit on top of the scan, not inside it. Rescan the area that moved and your library stays intact.

3 selection tools

Manual or AI-assisted
Isolate a single valve or a whole conveyor in a few clicks.

Your own asset types

Fields defined once, reused everywhere
Pull live values from SAP, Maximo, AVEVA PI and other ERP, MES, CMMS or IIoT systems.

Rescan-proof tags

Assets persist across updates
Rescan the area that changed. Skip the retagging.

1. Compose the Twin from your reality capture

Upload what you have: point clouds, meshes, 360 video, photogrammetry, CAD or IFC files, Gaussian splats. Each upload is processed into a data bundle, and the Composer stacks bundles as layers.

  • Shared registration: scans captured on the same survey targets align themselves when they land in the Composer. Anything else can be aligned by hand.
  • Layer order matters: the layer on top wins where two layers overlap, which is how a fresh scan replaces an old one.
  • Publish: the composition becomes the Twin, the single environment where assets live.

More on RealityComposer™

Screenshot of Prevu3ds Realitycomposer™ Interface Showing a Tls Dataset in Green and a Slam Dataset in Red Layered Together in a Unified 3d Environment Prevu3d

2. Draw the asset with the tool that fits the shape

  • Bounding box: one box or several boxes per asset. The most predictable option, and the one to teach first.
  • Magic Wand: click once and it computes the region it believes belongs together. Refine the box afterwards. It does best on well-separated equipment and is still being improved, so check its edges on tightly packed lines.
  • Pipe Selection: click along the run from node to node and corner to corner, set the diameter, confirm. Built for piping, conduit and other linear assets.
3d Scan Interface Showing Isolated Industrial Pump with Extraction Parameters Prevu3d

3. Give it an asset type and its metadata

An asset type is a template: preset sections and fields such as ID, specifications, health or usage. Pick the type and the fields appear, ready to fill, identical across every site that uses it.

  • Attach what the field team needs: PDFs, images, and link fields that point to records in other systems.
  • Talk in context: comments sit on the asset, and @mentions notify colleagues.
  • Points of interest versus assets: a point of interest is a pin in space with notes. An asset is equipment with geometry, which you can connect, move and export.
Screenshot of Realityplatform Asset Type Settings Showing Configurable Fields and Integration Link Options for a Forced Draft Burner Prevu3d

4. Organize assets your way, not ours

RealityPlatform™ does not force an asset hierarchy on you. An asset can be a valve, a conveyor line or an entire terminal.

Folders and spatial zones handle the rest, for example one site per factory and one zone per building, mirroring the structure already in your CMMS or ERP.

Prevu3d Interface Showing Metadata Tagging and 3d Model Segmentation of Mechanical Fire Pump System Prevu3d

Visual Asset Management with Semantic Layer — FAQs

It means managing equipment records through a 3D view of the real site instead of spreadsheets and floor plans. In RealityPlatform™, each piece of equipment in the scan becomes a tagged asset linked to its data in your CMMS, ERP or historian.

Draw a bounding box around it, click it with the Magic Wand, or trace it with Pipe Selection, then assign an asset type. The type brings its fields, which you fill by hand or link to a source system.

Yes, through custom HTTP integrations configured with the source system’s API. AVEVA PI has a dedicated integration. The asset ID stored in RealityPlatform™ is what links the two.

Yes. Assets are a layer above the reality data. Layer the new scan on top in the Composer and the assets, metadata and comments remain.

Yes. The RealityConnect™ plugin for Revit streams assets in as native families. Inventor, Siemens NX, Process Simulate, Plant Simulation and NVIDIA Omniverse have plugins too, and OBJ or STEP export covers other tools.

No. It accelerates selection one asset at a time. Automated segmentation of whole environments is in development.

Kinetic Vision Logo with Modern Black Typography and Three Colored Dots Prevu3d
Prevu3D is the connector that enables us to make reality capture data viable for integration into Omniverse. It's the starting point for segmenting and porting assets into simulation environments.
Alexander Doukas
Group Manager, Kinetic Vision

Visual asset management for the equipment you already own, tied to the data you already keep.

Two Construction Workers in Hard Hats and Hi vis Vests Discuss Plans One Holds a Blueprint and a Selfie Stick the Other a Clipboard Prevu3d
360 Camera vs Laser Scanner: When to Grab the Camera, and When to Call in the Crew
Learn more