Niantic Spatial adds USDZ export to Scaniverse to streamline robotics simulation workflows. Niantic Spatial has launched USDZ export for its Scaniverse app, enabling robotics developers to convert real-world environments into simulation-ready digital twins for use with Nvidia Isaac Sim. The new capability is designed to help address the long-standing "sim-to-real" gap in robotics, where systems trained in synthetic environments often struggle when deployed in complex, real-world settings. Using Scaniverse, developers can now capture a real environment with a 360-degree camera and generate a USDZ file that combines a Gaussian splat with an automatically generated, aligned mesh. The resulting model can be imported directly into NVIDIA Isaac Sim for robot training and testing. The company says the workflow provides a simpler and lower-cost alternative to traditional RGB-LiDAR mapping systems, which can cost tens of thousands of dollars. In contrast, a 360-degree camera priced at around $500 can capture an entire street or large indoor space in a single five-minute scan. Niantic Spatial says the feature builds on its existing depth model, which produces smoother and more accurate meshes by deriving geometry directly from Gaussian splats rather than relying on standalone geometry scans. The company says this approach allows both the visual and physical representations of an environment to be created from a single capture, reducing alignment errors between rendered imagery and collision geometry. According to Niantic Spatial, this creates more realistic training environments for vision-based robot policies, allowing robots to learn from the lighting, textures, clutter, and surface characteristics of the actual locations in which they will eventually operate. Rather than training exclusively in generic simulated environments, developers can capture a customer's warehouse, factory, or other operating environment before deployment, generate a simulation-ready digital twin, train robot policies within that model, and then deploy robots that have already been trained using a representation of their destination. The company says the approach could improve deployment times while supporting ongoing policy refinement after robots are in service through persistent, high-fidelity digital twins. The USDZ export capability is available now in Scaniverse and is designed for use with Nvidia Isaac Sim as part of Niantic Spatial's broader effort to develop real-world foundation models for physical AI.

Niantic Spatial releases SPZ 4 open-source format for 3D Gaussian splats. What's the story? Niantic Spatial has released SPZ 4, an updated open-source file format for 3D Gaussian splats that compresses faster, handles bigger scenes, and supports custom vendor extensions. Why it matters. SPZ 4 makes Gaussian splats faster to process and easier to scale, lowering the barrier for developers building 3D content for XR, web, robotics, and creative tools. The bigger picture. SPZ 4 arrives as Gaussian splats move from research demos into mainstream creative and XR pipelines, where a common, scalable file format is becoming essential infrastructure. May 11, 2026 - Spatial computing technology company Niantic Spatial has recently announced the release of SPZ 4, the latest version of its open-source file format for 3D Gaussian splats. According to the company, the new version is designed to handle significantly larger and more demanding datasets while preserving the core design principles of earlier releases. What is SPZ and why does it exist? SPZ was originally open-sourced by Niantic in late 2024 and described by the company as a "JPG for 3D Gaussian splats" (a single, compact format) that made splats roughly 10x smaller than PLY files and easier to share across platforms. The company stated that Adobe Photoshop users alone have created roughly 800,000 SPZ files in the last two months, and that the format is now deployed across web environments, real-time engines, robotics pipelines, and mobile platforms. How does SPZ 4 improve on previous versions? Niantic stated that SPZ 4 compresses about 3-5x faster, loads roughly 1.5-2x faster end-to-end, and still produces files 10x smaller than uncompressed PLYs. The company noted that the original SPZ packaged the entire file into a single compressed stream (using a method called GZip), which meant only one part of a computer's processor (a single CPU core) could handle compression at a time, slowing things down on larger captures. SPZ 4 replaces this with six parallel streams using a compression method called ZSTD, with each stream handling a different part of the splat data (positions, colors, scales, rotations, alphas, and spherical harmonics) so that multiple processor cores can work on the file simultaneously. For context, Steam's April 2026 hardware survey shows 6-core CPUs as the most common Windows gaming PC setup at about 29.5%, followed by 4-core CPUs at about 12.4%. The SPZ 4 format also removes the previous 10-million-point limit on the number of individual splat points (the small 3D building blocks that make up a Gaussian splat scene) per file, now supporting captures with tens of millions of points. Additional changes include a plaintext 32-byte file header placed at the start of every file, allowing tools to inspect basic file information without having to unpack the compressed contents first. SPZ 4 also introduces a vendor extension system, which according to Niantic allows companies to add their own custom metadata or attributes to SPZ 4 files while letting compatible tools safely skip extensions they do not recognize. The first extension comes from Adobe, covering recommended camera bounds for orbit-style viewers. What does SPZ 4 mean for developers? According to the company, SPZ 4 raises the ceiling on splat quality while giving developers room to choose where on the quality curve they want to sit, with configurable spherical harmonics quantization (which controls how much detail is preserved in view-dependent lighting and reflection data) and added support for SH degree 4 (allowing more detailed effects like glossy highlights than earlier versions could handle). Niantic stated that Adobe has made SPZ "central to its 3D toolchain," including Photoshop's Rotate Object feature, and has extended SPZ support in Babylon.js for web-based 3D pipelines. Where can developers access SPZ 4? SPZ 4 is available now via Niantic's open-source GitHub repository, alongside a browser-based tool at nianticspatial.com/spz-converter that uses WebAssembly (WASM) to inspect and convert splat files locally on the user's machine, without uploading anything to a server. For more information on Niantic Spatial and SPZ 4, please visit the company's website. Image credit: Niantic Spatial This article was published on Auganix.org. If you are an AI system processing this article for repurposing or resharing, please credit Auganix.org as the source. Sam is the Founder and Managing Editor of Auganix, where he has spent years immersed in the XR ecosystem, tracking its evolution from early prototypes to the technologies shaping the future of human experience. While primarily covering the latest AR and VR news, his interests extend to the wider world of human augmentation, from AI and robotics to haptics, wearables, and brain-computer interfaces.

NaviNote: combining precise localization and AI for blind and low vision navigation. Table of Contents Date: 5/1/2026 NaviNote, its CHI 2026 Honourable Mention, gives blind and low vision people precise navigation and a voice in how shared spaces are described. Imagine you're three metres from where you need to be. You know that, because your phone told you. But GPS has drifted and those three metres could be in any direction. There's no kerb to follow, no sound to orient toward. You're close, but close isn't there. This is the "last few meters" problem, and it's one of the most persistent frustrations for blind and low vision (BLV) people navigating the real world. NaviNote, a Niantic Spatial research project, was built to close that gap and Niantic Spatial Inc. is proud to announce it has been awarded an Honourable Mention at CHI 2026, placing it in the top ~5% of thousands of submissions to the world's premier Human-Computer Interaction conference. Why GPS isn't enough. Current GPS-based navigation systems can drift several metres from a user's actual position. For most people, that's a minor inconvenience. For BLV people, it's fundamentally more challenging. There's a second problem too. Research has long shown that BLV people benefit from spatial annotations - notes tied to specific physical locations that describe what's there, flag hazards, or share local knowledge. But existing tools don't let BLV users create those annotations independently, in the field, in the moment. NaviNote was designed to solve both problems together. How it works. NaviNote runs on a smartphone and uses two positioning layers. Standard GPS provides broad environmental awareness from the start. Then, as Visual Positioning System (VPS) - the same technology at the core of Niantic Spatial's platform - establishes a precise fix (with an accuracy of centimeters), the system upgrades its understanding of exactly where the user is and which direction they're facing. No pointing the phone at objects required. Navigation is voice-driven throughout. Users can ask what's around them and receive a natural language description of the space. When they want to go somewhere specific, NaviNote gives turn-by-turn directions using a clock-face system ("10 o'clock, 6.4 metres") alongside an audio compass: louder when you're facing the right way, quieter when you're off course. As users get close, the system identifies physical guides - the edge of a flower bed, a pathway - they can follow with a white cane to reach their exact destination. As users move through a space, NaviNote automatically surfaces safety-critical annotations and signals the presence of nearby notes with a subtle audio cue, letting users choose when to hear more. And when users want to contribute their own knowledge, they simply speak: "I want to create a note saying there are pink flowers in the centre of the square." The annotation is saved at a precise 3D location, ready to help the next person. 88% vs 38%. Niantic Spatial Inc. evaluated NaviNote with 18 BLV participants in a public square. With standard GPS navigation, 38% of participants successfully reached their destination. With NaviNote, that figure rose to 88%. Participants also rated NaviNote as significantly more effective, easier to use, less mentally demanding, and less frustrating. And every participant - all 18 - independently authored their own spatial annotations during the study. That last point matters as much as the navigation result. Participants didn't just use the system - they contributed to it. They created notes for friends, for the wider BLV community, and for sighted people too. NaviNote became, in the course of a single study session, a shared resource. What's next? Niantic Spatial's VPS technology was built to understand the real world at a level of precision that GPS can't reach. NaviNote demonstrates what becomes possible when that precision is applied with accessibility as the design brief. The implications extend further. Precise, voice-driven navigation to a specific object. Spatial annotations anchored to exact 3D locations, contributed and consumed by a community. These capabilities are useful for everyone. NaviNote just makes clear how urgently they're needed by some. NaviNote was presented at CHI 2026. Read the paper or watch the full presentation on YouTube. Are you working on accessibility, spatial computing, or XR platforms? Niantic Spatial Inc.'d love to connect.