Heading to United States Geospatial Intelligence Foundation (USGIF) #GEOINT2026? Stop by the Niantic Spatial booth to see how we turn multi-source data into a unified, centimeter-accurate 3D coordinate system, enabling precise geolocation and real-world understanding where GPS falls short. 📍Exhibiton Hall, Booth #803 📱Hands-on demos 🎁 Exclusive, free swag 🗓️ Book 1:1 meetings https://hubs.ly/Q04dD9Wj0 Interested in attending the event? Use "USGIFTEN26" for a 10% discounted rate. #NianticSpatial #GEOINT #USGIF #GeospatialIntelligence #PhysicalAI #Scaniverse #VPS #AI #DefenseTech #GPSDenied #DDIL
Niantic Spatial, Inc.
Software Development
Building a living model of the real world for machines
About us
We are an AI company building the foundational models for the era of robotics and spatial computing. Today's most powerful AI systems excel at language and code. But the real world isn't text — it's physical, dynamic, and constantly changing. If AI is to operate in the physical world, it needs a spatial brain built on the real world, not an imagined one. We are building that brain. Built on 25 years at the frontier of spatial computing. Niantic Spatial's founding team built Keyhole — the satellite imagery platform acquired by Google that they grew into Google Earth and Google Maps. They went on to create some of the largest location-based games ever made, and to pioneer Visual Positioning Systems at global scale. That foundation in real-world spatial data, geospatial AI research, and planetary-scale deployment is what Niantic Spatial is built on. Today we are an independent AI company deploying spatial intelligence across robotics, energy, construction, logistics, large venues, and the public sector.
- Website
-
https://nianticspatial.com/
External link for Niantic Spatial, Inc.
- Industry
- Software Development
- Company size
- 201-500 employees
- Type
- Privately Held
- Specialties
- Spatial Computing, Spatial Mapping, Augmented Reality, Virtual Reality, Extended Reality, Computer Vision, Visual Positioning System, Artificial Intelligence, Physical AI, Robotics, Geospatial Intelligence, and Spatial Intelligence
Employees at Niantic Spatial, Inc.
Updates
-
We’re heading to GEOINT 2026 hosted by the United States Geospatial Intelligence Foundation (USGIF) 💫 Visit our booth to see how we solve real-world mission operations—from capturing high-fidelity 3D models with Scaniverse to multisensor data fusion across ground and aerial inputs. See how our Visual Positioning System (VPS) enables precise ground-to-aerial localization in GPS-denied conditions, along with the ability to search, analyze, and annotate 3D environments. 📍 Experience the demos 🤝 Connect on geospatial intelligence 📅 Book a meeting: https://hubs.ly/Q04dh8wT0 #NianticSpatial #GEOINT #USGIF #GeospatialIntelligence #PhysicalAI #Scaniverse #VPS #AI #DefenseTech
-
Niantic Spatial, Inc. reposted this
My colleague David Hollin recently built a demo that simulates how robots can localize where they are in the real world while navigating. We took our robot to the top of Mount Tamalpais to show this off. This high-fidelity Gaussian splat was captured in just a few minutes with an Insta 360 X5 camera. Create your own Splats today with Sacinverse at https://lnkd.in/gmTE8xA8 Powered by: Niantic Spatial, Inc. Try this demo yourself: https://lnkd.in/gYEZXfNQ
-
Capture → Reconstruct → Localize Scaniverse enables high-fidelity 3D capture of the real world, creating the foundational inputs needed for reconstruction, localization, and semantic understanding. By turning physical environments into structured data, it unlocks a new layer of spatial intelligence that can be used across applications, workflows, and systems. Go deeper with our product team as they walk through the full journey. Watch the full YouTube playlist: https://hubs.ly/Q04d65QN0 #NianticSpatial #Scaniverse #Reconstruction #VPS #GeospatialAI #AI #PhysicalAI
-
Large venues aren’t static environments. They’re constantly evolving — events, layouts, equipment, and foot traffic change daily. Yet most teams are still working off outdated maps, fragmented data, or manual site walks. With Scaniverse, teams can quickly scan, capture, and reconstruct large venues using just a mobile device — turning physical spaces into high-fidelity, georeferenced 3D models. 🏟 Event & Operations Planning: Capture a stadium, convention center, or festival site and simulate layouts. 🦺 Safety & Emergency Readiness: Create up-to-date spatial models for evacuation planning, hazard mapping, and first responder coordination. 🛠 Facilities & Asset Management: Track infrastructure, monitor changes over time, and maintain a living record of the venue for maintenance and upgrades. Download and try it for yourself: https://hubs.ly/Q04cYk150 #NianticSpatial #Scaniverse #GeospatialAI #PhysicalAI #AI #3D #DigitalTwins #Reconstruction
-
Huge congrats to the team on an Honorable Mention at CHI 2026 🏆 — proud to see work like NaviNote pushing accessibility forward with spatial audio + precise navigation. #NianticSpatial #GeospatialAI #PhysicalAI #AI #XR #VPS #CHI2026
I’m excited to share that our paper, NaviNote, received the 🏆𝗛𝗼𝗻𝗼𝗿𝗮𝗯𝗹𝗲 𝗠𝗲𝗻𝘁𝗶𝗼𝗻 𝗔𝘄𝗮𝗿𝗱 at ACM CHI Conference 2026! NaviNote is a last-few-meters navigation and spatial annotation system for people who are blind or low vision (BLV). Combining visual positioning system (VPS)—a technology that determines users’ location with sub-meter-level accuracy—and agentic AI, NaviNote empowers BLV users to explore their surroundings by accessing and authoring precise audio notes for diverse audiences and purposes: -- Curious about a statue? Ask NaviNote and hear what others have said about it. -- Found an attractive landmark? Leave a note to record and share your experience. -- Meeting a friend? Leave a note at the meeting place, and your friend can ask NaviNote to provide turn-by-turn instructions to the meeting place. -- Notice an inaccessible spot? Leave a warning to alert others and to suggest improvements. We ran two studies with BLV participants. In a formative study (N=24), we identified six categories of desired spatial annotations and learned that precise navigation is a prerequisite for exploration. In our evaluation (N=18), NaviNote significantly improved navigation success, helped participants understand their surroundings, and enabled all 18 to independently create annotations. This project was done during the internship that Yuheng W. and I did at Niantic Spatial, Inc., in collaboration with University College London. Huge thanks to our advisors, Jessica Van Brummelen and Gabriel Brostow, for their guidance throughout this project, and to our wonderful collaborators, Charlie Houseago, Filipe Gaspar, Filippo Aleotti, Dorian Galvez-Lopez, Oliver Johnston, Diego Mazala, Guillermo Garcia-Hernando, and Maryam Bandukda, PhD, for their support and collaboration! I won’t be in Barcelona in person, but Jess will be presenting the work. Looking forward to the presentation! Paper: https://lnkd.in/gxsBcbqt Video presentation: https://lnkd.in/gCmnF5TX
-
-
Niantic Spatial, Inc. reposted this
With Scaniverse you can also use 360° drones to capture large area scans and process those through our 360° pipeline to generate high-fidelity gaussian splats. This was captured on an Antigravity A1 drone. Powered by: Niantic Spatial, Inc. Try out Scaniverse now: https://lnkd.in/gmTE8xA8
-
Unlike traditional GPS, which can drift several meters and fails indoors or in dense environments, Niantic Spatial’s Visual Positioning System (VPS) matches live visual data against a geo-referenced spatial model to return a highly accurate 6DoF pose. In this deep dive, Senior Product Manager, Eugene Chong, explores VPS in action using camera input and machine learning to determine position and orientation relative to the physical world. This enables: 🔹Precise positioning and orientation (~10-15cm accuracy in mapped areas) 🔹Robust performance in GPS-denied or degraded environments 🔹Persistent spatial anchoring across devices and sessions From localization to a shared coordinate system for machines—built from continuously updated scans, imagery, and sensor fusion—this is the foundation for systems across robotics, autonomy, and infrastructure to understand and act in the real world, not just simulate it. Watch the full video on YouTube: https://hubs.ly/Q04cs_lQ0 #NianticSpatial #Scaniverse #VPS #GIS #GeospatialAI #GEOINT #PhysicalAI #AI #Robotics
-
A new NSDK 4.0 update is live. So what? It means building with real-world spatial data just got a lot easier. At the same time, we launched Scaniverse last week. So how do they fit together? Scaniverse creates the world → NSDK lets your apps understand and interact with it. You can: 🔸 Capture and process real-world spaces with Scaniverse 🔸 Bring those assets into NSDK for localization and spatial computing Or, if you’re building your own app: 🔸Integrate NSDK directly and use features like VPS2 out of the box. It’s a connected pipeline: Capture → Reconstruct → Localize → Build → Deploy Develop with Unity, Swift, Kotlin, and early ROS2 support — and go from real-world data to real-world applications faster than ever. 🔗Get started: https://hubs.ly/Q04chT6h0 #NianticSpatial #Scaniverse #NSDK #GeospatialAI #PhysicalAI #XR #Robotics #Unity #Kotlin #Swift
-
NSDK 4.0 updates are live 🚀 This release builds on the foundation led by Ken Wolfe and is a direct reflection of developer feedback, with updates designed to make building faster, simpler, and more intuitive: 🔹 Reworked Swift & Kotlin SDKs (native patterns, modern workflows) 🔹Faster setup with Xcode & Android Studio integration 🔹New “Getting Started” docs for quicker onboarding 🔹VPS 2.0 for more reliable real-world localization—even in unmapped spaces 🔹On-device playback for easier testing and iteration 🔹Simplified authorization flow Get started here: https://hubs.ly/Q04cgLTl0 #NianticSpatial #NSDK #SDK #XR #AR #GeospatialAI #PhysicalAI #Unity #Kotlin #Swift
-