Niantic Spatial 🌎

We're building real-world foundational models for physical AI, so people and machines can work efficiently and safely together.

[Niantic Spatial 🌎 @NianticSpatial]

Introducing object interaction for real-to-sim training.

Identify, extract, move, or remove objects in captured environments to test navigation, obstacle avoidance, and baseline interactions.

Try scene variation in beta with up to 10 scenes: https://t.co/wgiq9odsHl

#EmbodiedAI #Robotics #PhysicalAI #AI #RealtoSim

[asim ᯅ @asimahmed]

today at @nianticspatial we're introducing object interaction (beta) in gaussian splats for teams working on embodied AI.

this opens up new possibilities for domain randomization via scene variation for real2sim training anchored in our state of the art real-world reconstructions.

[Niantic Spatial 🌎 @NianticSpatial]

@asimahmed 👏

[Justin Ryan @justinryanio]

Gaussian splats are just so freaking cool.

I captured Niantic Spatial’s lobby in under five minutes with my Insta360 X6, processed it in Scaniverse, and now I’m exploring it on Apple Vision Pro.

Nearly 3.9 million splats. The detail is incredible.

[Niantic Spatial 🌎 @NianticSpatial]

We’re building a 4D model of the world so machines can reason, learn, and act.

The foundation: photorealistic 3D reconstruction with accurate, metric-scale geometry for field operations, remote planning, robot policy training, and more. 

These possibilities are endless when real-world places become the foundation of where you build.
https://t.co/kxIUEa4VQg

#3DReconstruction #GaussianSplatting #SpatialAI #EmbodiedAI #Robotics #RealtoSim

[Niantic Spatial 🌎 @NianticSpatial]

Give robots a faithful reconstruction of their world.

Niantic Spatial’s depth-regularized reconstruction unifies photorealism and reliable geometry for meshes, collision surfaces, and robot training.

Learn more about the state-of-the-art geometry for autonomy: https://t.co/yfBGFwmttE

#Robotics #PhysicalAI #EmbodiedAI #AI #GaussianSplat #3DReconstruction #NeuralRendering

[Niantic Spatial 🌎 @NianticSpatial]

Are your robots ready for the real world?

Start with Niantic Spatial’s Embodied AI Places Library: 100 off-the-shelf warehouse, residential, and outdoor environments for real-to-sim training.

https://t.co/ojxmwDsuMM

#Robotics #EmbodiedAI #RealToSim #3DReconstruction #GaussianSplatting #3DGS #IsaacSim #AI

[Niantic Spatial 🌎 @NianticSpatial]

More accurate depth and geometry for robot training.

Niantic Spatial uses depth evidence to guide Gaussian splat optimization. By comparing rendered splat depth against per-view depth estimates, we encourage Gaussians to occupy the right locations and surfaces to remain consistent across viewpoints.

In our comparisons, this reduces noise in extracted surfaces, particularly across large planes and low-texture regions, combining photorealistic rendering with reliable geometry in one representation.

Learn more: https://t.co/EncUs6sDnx

#EmbodiedAI #PhysicalAI #AI #Robotics #GaussianSplat #3DReconstruction #RealtoSim

[Justin Ryan @justinryanio]

I’ve been using Niantic Spatial for months to create Gaussian splats, so it was really cool to meet the team and check out their office with a hidden doorway.

Thanks for having me over @asimahmed!

[Niantic Spatial 🌎 @NianticSpatial]

We’re creating a 4D living model of the real world.

From indoor spaces to entire cities, Niantic Spatial is turning the physical world into accurate, high-resolution 3D models.

Captured by Spexi. Reconstructed by Niantic Spatial.

Explore what’s possible: https://t.co/TG7GO7d3Xp

#Scaniverse #3DReconstruction #SpatialAI #GaussianSplat #DigitalTwins #DroneImagery #3DGS #AI

[Niantic Spatial 🌎 @NianticSpatial]

What happens when AI gains a deeper understanding of the physical world?

Niantic Spatial CEO, @InhiChoSuh will explore spatial intelligence and world models in “New Frontiers of Competition” at The Great Redesign at Yale.

🗓️ November 4 | 10:45–11:45 AM

Register here: https://t.co/Q7ePUODntv

#EnterpriseAI #PhysicalAI #SpatialAI #AI

[Niantic Spatial 🌎 @NianticSpatial]

Train policies on geometry robots can trust.

Our Places Library pairs a Gaussian splat with an aligned collision mesh, preserving the planes, edges and clearances that matter for sim-to-real transfer.

Download a free sample: https://t.co/It4Xufbj3K

#EmbodiedAI #PhysicalAI #AI #Robotics #GaussianSplatting #3DReconstruction

[Niantic Spatial 🌎 @NianticSpatial]

At #AES2026, we combined scene search, 3D reconstruction and visual positioning in one remote spatial collaboration workflow, helping enterprises extend remote expertise, reduce time spent finding work and turn reality-capture data into action.

Learn how we’re reshaping Enterprise AI for Oil & Gas: https://t.co/ZTfTFvMnyK

#SpatialAI #EnterpriseAI #PhysicalAI #AI #3DReconstruction #ComputerVision #VisualPositioning #VPS

[Niantic Spatial 🌎 @NianticSpatial]

How do machines understand where they are, and what’s around them?

Niantic Spatial CTO Brian McClendon joins @jonahbliss of @CurbivoreNews to discuss mapping, visual positioning, and the real-world context #PhysicalAI systems need to operate.

Full podcast here: https://t.co/HdVY2AyNOH

#GeospatialAI #SpatialAI #AI #Robotics

[ssılqɥɐuoɾ @jonahbliss]

How do AVs understand the world? I was delighted to chat with geospatial legend Brian McClendon, CTO of @NianticSpatial, about the future of location intelligence.

[asim ᯅ @asimahmed]

a scene that looks good is the easy bar. state-of-the-art geometry a policy can train against is the hard one.

warehouse from our places library: high-fidelity 3d gaussian splat + aligned mesh, collider-ready for robot simulation.

powered by @nianticspatial

[Niantic Spatial 🌎 @NianticSpatial]

Photorealism is not geometry. A splat can look perfect from the camera and still be warped underneath.

State-of-the-art geometry means surfaces stay where they belong, from a sparse phone capture up to a full facility. 

Learn how we got there: https://t.co/DJTNzLaj1V

[Niantic Spatial 🌎 @NianticSpatial]

✅ #AES2026 Day 1 is a wrap!

Thank you to everyone who attended our opening keynote where CEO Inhi Cho Suh shared our vision for Enterprise AI in the real world.

Come visit us at Booth #120, to experience our Intelligent Field Operations demo and learn how our 3D reconstruction technology supports applications ranging from city-scale digital twin simulations to policy training for embodied AI.

Connect with our team and book a meeting at AES: https://t.co/hJzc7LWG5q

#PhysicalAI #SpatialAI #EnterpriseAI

[Niantic Spatial 🌎 @NianticSpatial]

Are you a student or recent grad interested in Spatial AI and XR?

Niantic Spatial is proud to sponsor MOLTEN, a one-day hackathon for prototyping the tools of spatial creation during #LATechWeek.

Participants get one month of Scaniverse Pro to capture real-world spaces and turn them into 3D assets for XR prototypes.

Ready to build? RSVP: https://t.co/TAKo0RNhZG

#LATechWeek #SpatialAI #PhysicalAI #AI #3DReconstruction #XR

[Niantic Spatial 🌎 @NianticSpatial]

AI needs more than intelligence to operate in the physical world. It needs ground truth.

Niantic Spatial’s Inhi Cho Suh and Eugene Chong join the season finale of Zero-Shot Learning to discuss spatial data, safer AI behavior and real-to-sim training.

Watch: https://t.co/lZd8VONmMD

#SpatialAI #PhysicalAI #AI #WorldModels #Robotics #EmbodiedAI