Good morning. ☕ Quick one for you. Satellites can see your roof. Mapping cars can see the road they're allowed to drive on. But neither one can see a stairwell, a loading bay, a busy market street or the inside of a warehouse. That's the gap @vangrid_io is filling and it's a bigger deal than it sounds. They're building the spatial data layer for Physical AI. Real-world ground truth. And the wild part is how they do it: they turn ordinary smartphones into a decentralized perception grid. Here's the flow: Contributors record short videos on their phones faces and plates blurred on-device, so privacy comes first. Those videos get reconstructed into 3D models point clouds, Gaussian splats, that whole world. Every capture gets fingerprinted content hash + rough location + timestamp Merkle-batched and anchored on Base via EAS. Buyers can verify provenance on-chain. No trust-me-bro. Organizations and AI agents post bounties fund them in USDC (Base or Arc via x402) and get the verified reconstruction back. An agent can literally hire a human with one HTTP call no account no API key. The numbers aren't small either: 1.25M+ captures hundreds of thousands of users thousands of attestations real USDC settled. And they raised a $9M seed from HashKey https://t.co/NCPq6PrmOb Capital Animoca Brands and others. Who's this for? Embodied AI, robotics, world models autonomous logistics critical infrastructure sovereign defense. Basically anything that needs to understand the physical world at ground level. The grid is already live capture at https://t.co/1NAY9sekE9 or dig through the on-chain activity at https://t.co/S9XQdWRL3Y. Physical AI can't run on guesswork. It needs continuous ground-level verifiable spatial data. This is one of the clearer DePIN plays to supply exactly that. Have a good one. Go capture something. 📈