Pioneering in Spatial AI
Enter
Reshaping the discrete world into a digital continuum with World-Engine.
Overview
Zeno is a comprehensive platform designed to create a persistent, shared digital layer over the physical world. Our mission is to establish a set of models and a robust system to describe and connect the physical world with dynamic virtual worlds.
This platform offers a business model of next-generation interaction, which could utilize and digitize the space of the physical world. There are many ideas such as AR experiences, robots or AI agents tasks, digital ads, space monetization etc.
Scroll Down
Real World Understanding
Spatial Intellgence
High precision fused visual positioning algorithm and vision world model for the platform to determine a pose of an agent in the real world, with the agent's camera and other sensor signals.
Blended Universe for
Humans, Robots & AI Agents
By digitizing physical spaces and overlaying multiple parallel virtual worlds, we offer a blended universe for real world agents, including human beings, robots, smart devices, and virtual world agents, including human controlled avatars, AI-powered NPCs, and other bot-like programs to live together.
Core
Concepts
01
Real World,
Virtual Worlds,
Spatial Anchors
One physical world we inhabit as both human or robots with multi-digital worlds created on top of the digitally-mapped real world in phenomenal zones. Multiple, separate virtual worlds can coexist and overlap with the same physical coordinates. Similar to lighthouses on a dark ocean, the space anchors work like a torch that lights up the surroundings. And multiple anchors construct a web of anchors layer that bridges the physical and digital worlds, supporting spatial discovery and access controls.
02
Agents
An entity object lives in a virtual world, and may or may not have its representation in the real world. It can represent a human, human-controlled avatar, robot, AI-powered NPC, or anything that can actively be somewhere and able to do something like observing, moving or interacting with the environment and other agents in the blended worlds.
03
World Engine
Drives a virtual world and runs all agents (including human driven, robot driven or NPCs) follow certain rules in the world engine and are involved or do their designated tasks by themselves. Like the physical world that is ruled by physics laws, the virtual worlds are driven by virtual world engines with alternative sets of rules, similar to reality or purely magical.
04
The Zeno World
The unique public virtual world for the entire earth, called “The Zeno World”, 1:1 mapping the public area of the physical earth run by world engine. This world is run by the platform itself, to provide an arena for everyone on earth to digitally inhabitate and play. They could place and consume content, interact with agents, like robots, NPCs, AI-bots. Some users are also content creators or developers but they can interact directly in this public open world.
Architecture
latest News
November 10, 2025
Famed AI specialist Dr. Fei-Fei Li detailed why the next breakthrough in AI will come from spatial intelligence, or systems that can understand, reason about, and generate 3D, physics-consistent worlds.
June 4, 2025
A16Z and Godmother of AI, Fei-Fei Li explains why spatial intelligence could reshape the future of robotics, creativity, and computational interfaces.
August 19, 2021
Tesla develops and deploys autonomy at scale in vehicles and robots, believing that advanced AI for vision and planning is the only way to achieve a general solution for full self-driving, bi-pedal robotics and beyond.
December 5, 2024
A16Z believes neural networks that simulate visual and physical environments are transforming robotics by replacing traditional simulators with generative models that can predict and recreate real-world sensory outcomes.
December 29, 2024
Nvidia bets Big on Robots. The chipmaker laid out a vision for dominating so-called physical AI.
May 16, 2017
OpenAI created a robotics system by deploying physical AI, trained entirely in simulation and deployed on a physical robot, which can learn a new task after seeing it done once.