Remember the Pokémon GO craze in Cape Town began in October 2016, when the game officially launched in South Africa?
When Pokémon Go launched in South Africa in October 2016, players across Cape Town flocked to parks and promenades to chase virtual creatures through real-world landmarks. Niantic says images and AR scans gathered through Pokémon Go have produced a dataset exceeding 30 billion photographs now used to train visual navigation for delivery robots, reports Cape {town} Etc.
Cape {town} Etc Discount Alert! Unlock Cape Town’s best experiences for half the price! From unforgettable adventures to hidden gems and mouthwatering restaurants, these limited-time deals won’t last long. Snag your discount before they’re gone!
Pokémon Go players unknowingly created a rich, street-level archive that spans parks, shops and sidewalks, capturing multiple angles, times of day and weather conditions.
Niantic says the material underpins a Visual Positioning System that can locate a robot with centimetre accuracy and replace GPS in dense urban areas.
Coco Robotics has begun piloting delivery bots that use the system to find addresses more precisely, a move Niantic describes as the first practical use case of its spatial AI.
Observers and social posts reacted with surprise and humour. Mark Gadala-Maria wrote: ‘This is wild. 143 million people thought they were catching Pokémon.’
A short clip from NewsForce framed the phenomenon as a vast, unpaid data effort. The video notes ‘no V2, no benefits, not even a thank you email.’
POKÉMON GO PLAYERS TRAINED 30 BILLION IMAGE AI MAP
Niantic says photos and scans collected through Pokémon Go and its AR apps have produced a massive dataset of more than 30 billion real-world images.
The company is now using that data to power visual navigation for delivery… pic.twitter.com/FIs65uO3sx
— NewsForce (@Newsforce) March 15, 2026
It adds that the dataset lets a system ‘know exactly what it is within a few centimetres. The clip further says: ‘Your order arrives at your actual door and not at your neighbour’s house.’
Players’ images arrive with rich metadata, including location, facing direction and time, which turns casual scans into precise spatial models. Niantic says the map updates as robots feed back new imagery, creating a living twin of city streets.
Privacy advocates call for clearer consent and reuse policies while industry voices point to improved delivery accuracy.
The episode shows how everyday behaviour can quietly build powerful digital systems and why transparency and accountability matter as those systems move from games to sidewalks.
Local regulators and cities must scrutinise future deployments in public.
Be the first to know – Join our WhatsApp Channel for content worth tapping into! Click here to join!
Also read:
Introducing Miss AI: World’s first artificial intelligence beauty pageant
Picture: Pokémon GO / Facebook
