Spatial Insider
← Capture & Create3D mapping

MultiSet Adds 3D Scan Exports for NVIDIA’s Robot Simulator

The latest update also lets devices find nearby maps automatically, combines separate survey scans, and makes mapped spaces shareable in a browser.

· 2 min read

ShareXFacebookLinkedIn
MultiSet illustration showing industrial machinery as a purple point cloud on the left and solid geometry on the right
MultiSet’s illustration of an industrial space and its digital map.Image: MultiSet AI · Illustration

MultiSet (opens in a new tab), a platform that turns scanned places into maps for AR apps and robots, now exports those spaces for use in NVIDIA Isaac Sim and Omniverse.

Released September 19, version 2.3.3 (opens in a new tab) adds textured USDZ downloads for every map. The same capture can provide a scene for simulation and a map for positioning a device in the real space.

The update also removes the need for an app to name the map it wants to search. It can send GPS coordinates instead, and MultiSet selects a nearby map (opens in a new tab) from the account before matching camera images to find the device’s position and direction. The map must already have real-world coordinates and a heading.

Separate E57 laser scans can now be combined without overlapping imagery (opens in a new tab) if they share a survey coordinate system. MultiSet uses those surveyed positions to assemble a MapSet, and further scans can be added without moving existing ones.

New share links (opens in a new tab) let recipients explore a map or MapSet in a browser without an account. Links include the 3D mesh and an optional 360-degree tour. Individual maps can also include a photographic 3D view called a Gaussian splat.

A 3D scan of a brewery floor displayed in MultiSet’s browser viewer
A brewery scan in MultiSet’s browser viewer.Image: MultiSet AI

Owners can make links expire, limit them to one opening, or revoke access.

MultiSet says the release is live on its public cloud (opens in a new tab), with private-cloud, self-hosted, and on-device deployments following their own rollout schedules.

More from Capture & Create.

Browse all stories
Two hands reaching toward a blue, white, and yellow virtual block boat in Blockworks on Apple Vision Pro

Blockworks · 2 min read

Blockworks Turns Your Words Into Builds with AI on Apple Vision Pro

A new AI demo makes a boat, a bicycle picture, and a statue from virtual blocks you can pick up and take apart.

Justin’s 3D character with visible skeleton bones in Tripo Studio

Tripo AI · 4 min read

Tripo Adds 3D Animation From Text Prompts

Creators can generate short actions from prompts and combine up to five motion segments, with skeleton options for Unreal Engine, Unity, and avatar workflows.

NVIDIA’s Robo Olympics demonstration showing nine simulated robot sports in a grid

Robot simulation · 1 min read

NVIDIA’s Robo Olympics sends simulated robots sprinting, skating, and kickboxing

Sports videos and written prompts guided OpenAI’s Astra in building the events, with NVIDIA’s simulation tools turning motion references into tested robot movements.

Preparing the Spatial Insider studio