Niantic builds “geospatial” AI model based on Pokémon Go
Niantic builds “geospatial” AI model based on Pokémon Go


Nianticthe company behind phenomena such as Pokémon Go, announced its new ambitious project: the “Large Geospatial Model” (LGM). This initiative promises to redefine the way we perceive and interact with the environment through advanced artificial intelligence. In this article, we explore the details of this innovation that could transform our daily lives.
What is the large Geospatial model?
Niantic’s Large Geospatial Model combines millions of scans made by users of its games, including Pokémon Go And Ingress. By integrating this data, the company wants to enable computers and robots to understand and interact with the world in new ways. Victor Prisacariu, chief scientist at Niantic, emphasizes that this model is based on cutting-edge technology that unites 3D geometry and semantic understanding.
The foundations of the model
- Spatial intelligence: The LGM uses neural networks from Niantic’s Visual Positioning System (VPS).
- User scans: These millions of scans are made from players’ smartphones, which unwittingly share valuable information about the environment.
- Pedestrian perspective: The data collected represents places that cars cannot reach, providing an enriched and unique view.
Revolutionary technology
With the creation of the Grand Geospatial Model, Niantic is taking a significant step toward improving the interaction between humans and technology. Over the past five years, the company has focused its efforts on developing its visual positioning system. This system uses a single image from a smartphone to determine its position and orientation in a 3D map, designed with input from users around the world.
Practical applications of LGM
The applications of the Large Geospatial Model could be vast and varied, offering a multitude of possibilities in various fields:
- Video games: Enriching gaming environments with realistic elements, increasing the user experience.
- Robotics: Improved robot navigation in complex environments, making it easier for them to interact with the real world.
- Augmented reality: Integration of virtual elements into real landscapes for educational, tourism and commercial applications.
The challenges to be met
However, creating the Grand Geospatial Model does not come without challenges. Niantic must navigate data privacy and security issues while maximizing the usefulness of that information. Transparency will be key to convincing users to share their data with confidence.
A promising future
The Large Geospatial Model is a strong signal of the innovation underway at Niantic. By capitalizing on the contributions of its community and using advanced technologies, the company is redefining the boundaries of spatial understanding. The implications of this advancement are promising and could revolutionize our interaction with the world.
As this technology develops, it will be fascinating to follow its evolution and discover how it will inform our conception of augmented reality and interactive environments.
In conclusion, Niantic’s Large Geospatial Model represents much more than a simple technological project. It embodies a vision of the future where artificial intelligence and augmented reality intertwine to enrich our experience of the world. We look forward to the next steps in this geospatial revolution.






