Good morning. On your way out today, every building you pass is data a robot still does not have.
Physical AI has a simple problem. We mapped the outside world with satellites, but we have almost nothing for inside. A building, an alley, a room. Robots need that detail to move, and we do not have it.
Vangrid @vangrid_io took the practical route.
Most teams tried to fix mapping with expensive rigs. LiDAR backpacks, camera cars, hardware that cannot scale. Vangrid did the opposite. No new device. Just the phones in our pockets. The real advantage is not better hardware, it is wider reach.
I like the bounty setup too. AI agents put up USDC on Base or Arc for the exact place they need, people nearby capture it, then it turns into a 3D model, checked onchain with Merkle fingerprints and paid through API. Payment only happens after verification, so data quality is built in.
$9M from HashKey and Animoca and thousands of attested videos already. Early money plus early traction.
For me this is DePIN with actual use. Robots need to see inside our world, and this is the cheapest way to give them that.
If your phone can earn by capturing your city, what spot would you film first?
#PhysicalAI #Vangrid #Base