Pokémon Go AR Data Fuels Centimeter-Accurate Delivery Robot Navigation

Got story updates? Submit your updates here. ›

Niantic Spatial, an AI spinout formed in 2025, has turned years of mobile gaming data from Pokémon Go and Ingress into a high-precision world model of the physical environment. The company is now commercializing this technology through a visual positioning system that can locate devices to within a few centimeters using only camera input and map context. Coco Robotics, a last-mile delivery startup, is using Niantic Spatial's system to guide its fleet of sidewalk robots through dense urban areas where GPS alone is too unreliable.

Why it matters

As mapping has evolved from 2D to 3D and into dynamic "digital twin" simulations, the core link between map coordinates and physical locations has remained. What is changing is the primary consumer of those maps - increasingly, it is machines rather than humans. The same spatial intelligence that once kept virtual Pikachu aligned with the sidewalk is now being repurposed to keep delivery robots on course through traffic and weather.

The details

Niantic Spatial's visual positioning system (VPS) localizes devices based on what they see rather than relying on radio signals alone. The company has aggregated data from Pokémon Go and Ingress, which encouraged players to visit specific real-world locations. This produced a dense global dataset of images paired with metadata like latitude, longitude, camera orientation, and device pose. Niantic Spatial trained its models on roughly 30 billion images, allowing its system to infer an exact location and orientation from a handful of current images. Coco Robotics' delivery robots use this technology to fuse GPS with camera-based localization, improving reliability in dense urban areas where GPS is unreliable.

• Niantic Spatial was formed as an AI spinout in 2025.
• Pokémon Go and Ingress, the mobile games that provided the data, were released in the years prior to 2025.