Niantic uses Pokémon Go scans to train Coco delivery robots

Pokémon Go creator Niantic is repurposing years of augmented reality data to train autonomous food delivery robots for Coco Robotics. With that data, the robots can navigate complex urban environments without relying on traditional, often unreliable GPS signals.

For nearly a decade, millions of players have been scanning landmarks and streets to catch virtual creatures, which turned a game into one of the largest crowdsourced data collection efforts in history. Niantic is now applying this "spatial intelligence" to help Coco Robotics navigate the "urban canyons" of cities like Los Angeles and Chicago. In these environments, GPS signals often bounce off tall buildings, leading to significant navigation errors for traditional delivery systems.

A Visual Positioning System (VPS) lets these robots determine their location by looking at their surroundings instead of relying on satellites alone. They can move safely through chaotic city streets, avoid obstacles, and find the most efficient path to a customer's door. Food delivery moves from guessing where a courier is to knowing exactly how a robot is moving through a three dimensional space, with much less of the friction that typically slows down the final leg of the delivery journey.

Centimeter-level mapping resets delivery windows

Hyper-local mapping is gaining weight as a competitive factor. When robots can navigate with centimeter-level precision, the logic of delivery windows and logistics planning changes, and traditional maps fall short. Companies that rely on standard mapping tools may soon struggle to compete with the efficiency and predictability of AI-driven spatial fleets.

Autonomous fleets are spreading through major metropolitan areas and raising customer expectations for speed. Platforms can measure their performance against the market with Benchmark Delivery ETA, see where automation provides the biggest gains, and adjust their strategies accordingly.

If you want to see how robot fleets change delivery times in your cities, contact our team at DoubleData.

Source: https://www.polygon.com/pokemon-go-data-ai-robots-niantic/

Download our new report

Uber Eats x London [2025]

We analyzed venue coverage, quality distribution, promotional strategies, pricing thresholds, and logistics models across London to show what drives competitive advantage on Uber Eats. Our report is the first open-access, data-driven benchmark of Uber Eats' competitive strategy, written for food industry decision-makers.

Get access