A 30-billion-image dataset built by players over the last decade is now being used to train an AI navigation system ...

The cameras, advanced perception and reasoning software that enable robots to see are more than tools for navigation or task execution; they are the visual cortex for humanoids, allowing them to move, ...

Because the data was captured by players walking on foot, it provides a unique street-level perspective that satellite imagery cannot replicate. In 2020, Niantic introduced "AR Mapping" tasks, where ...

Pokémon Go's vast player-submitted image data is now powering real-world robotics. Niantic Spatial converts these billions of ground-level images into a photorealistic street-level model, enabling ...

Have you ever wondered how insects are able to go so far beyond their home and still find their way? The answer to this question is not only relevant to biology but also to making the AI for tiny, ...

One of the biggest obstacles to deploying very small autonomous robots is navigation. As machines shrink, the hardware required to determine position and direction does not scale down easily.