Pokémon Players Train Company's 30 Billion Photos to Build "AI World Model," Boosting Delivery Robot Industry

The globally popular augmented reality game Pokémon Go is developed by Niantic, whose AI company Niantic Spatial is now utilizing billions of city images captured by players over the years to build a “visual positioning system” and an AI world model capable of understanding the real world. This technology can accurately locate devices in urban environments with unstable GPS signals and has been tested in collaboration with delivery robot companies, opening new possibilities for navigation of robots and AI in real-world settings.

Pokémon Go Player Images as AI Training Data, City Photos Constructing World Models

Since its launch in 2016, Pokémon Go has quickly become a worldwide hit, with players capturing Pokémon in the real world through their phone cameras. Developed by Niantic, this well-known AR game continues to maintain over 100 million active players annually, even years after its release.

However, players constantly point their phones at city buildings and landmarks during gameplay, inadvertently accumulating a vast amount of image data.

Niantic’s AI company Niantic Spatial recently announced that it has collected and organized approximately 30 billion photos from urban environments worldwide, each with precise geographic and shooting information, such as phone orientation, movement speed, and camera angle. These data are now being used to train AI to build a “world model” that understands real-world space.

(Deep Dive: Are LLMs Flawed? Why Yang Likun’s AMI Focuses on the World Model Approach)

Visual Positioning System vs. GPS: AI Can Determine Precise Location Using Building Images

NewsForce reports that Niantic Spatial’s latest development is a Visual Positioning System (VPS). This AI model analyzes photos of buildings or landmarks to determine the user’s location with centimeter-level accuracy.

The company states that its database now covers over one million landmark locations worldwide. At each site, thousands of images taken at different times, angles, and weather conditions are stored. By comparing these visual features, AI can estimate the device’s position and viewing direction, providing highly accurate localization results.

Brian McClendon, CTO of Niantic Spatial, explains that this differs from traditional GPS, which relies on satellite signals. Instead, VPS uses “what it sees” to determine location:

In dense urban environments with tall buildings, GPS signals often drift, leading to errors of tens of meters or even wrong directions.

While such errors may not significantly impact general users, they can cause major issues for robots requiring precise navigation. Therefore, image-based positioning technology is a solution that robotics companies are paying close attention to.

From Catching Pokémon to Delivery: Delivery Robots Start Using Niantic’s Technology

Niantic Spatial has begun collaborating with Coco Robotics to test this technology. Coco has deployed about 1,000 delivery robots across several cities in the US and Europe, mainly for food and grocery delivery. These robots are roughly the size of small suitcases, capable of carrying up to eight large pizzas or four grocery bags.

The company states that, although the robots have completed over 500,000 deliveries, GPS inaccuracies sometimes make it difficult for them to stop precisely at restaurant or customer doorways:

With Niantic’s visual positioning model, robots can analyze their surroundings using four onboard cameras to more accurately determine their location and direction, improving delivery reliability.

The Era of Robots Is Coming: Niantic Aims to Create a “Living Map”

John Hanke, CEO of Niantic Spatial, explains that the initial goal of developing visual positioning technology was to support AR glasses and augmented reality applications. However, as the robotics industry rapidly develops, the company has shifted its focus toward robot navigation.

He states that the company is building a system called the “Living Map,” a highly detailed and continuously updated digital world model that adapts to changes in the real environment.

In the future, delivery robots, smart devices, and even AR headsets could serve as data sources, constantly transmitting environmental information to keep the digital world aligned with the dynamic real world.

AI Needs to Understand the Real World: “World Model” Becomes a New Tech Focus

In recent years, AI research has increasingly emphasized the concept of a “world model.” While large language models (LLMs) excel at processing text and knowledge, they still face significant limitations in understanding physical space and real-world environments.

By integrating maps, images, and environmental data, world models aim to enable AI to comprehend objects, spatial relationships, and environmental changes. Companies like Google DeepMind are also developing models capable of generating virtual worlds for training AI agents.

Niantic Spatial takes a different approach by using vast amounts of real-world image data to gradually reconstruct a digital model of the physical environment. As more data accumulates, this system could become a crucial infrastructure for future robots and AI to understand the real world.

This article Pokémon players train the company with 30 billion photos to build an “AI world model,” aiding the delivery robot industry first appeared on Chain News ABMedia.