Vangrid：用路人手机采集 3D 地点数据
需求方发布地点和 USDC 赏金，附近用户用手机绕场拍摄，经预览确认后生成 3D 模型，AI agent 可通过 API 下单并托管付款。
GM gVAN ☀️ Getting fresh data about a real place can take weeks. A mapping van needs to drive there. A sensor box needs to be built, shipped, and installed. But a phone? Someone nearby may already have one in their pocket. That is the idea behind @vangrid_io. It turns everyday phones into tools for collecting 3D data about real places. Here is how it works: → A requester posts a location and offers a USDC reward. → Someone nearby opens the app and walks around the area with their phone. → A preview checks the capture before the full 3D model is created. Faces and license plates are blurred on the device, while records are secured on Base so their origin can be verified. It gets more interesting with AI agents. An AI agent can request data through an API, pay for the job, and let a nearby person do the physical work. Funds stay in escrow until the submission is accepted, with refunds if the job is cancelled or expires. Why does this matter? Robots and AI need real world data to understand places they have never seen. Vangrid gives them another way to get that data without waiting for a mapping crew. The real world cannot be mapped by software alone. Sometimes, you need someone on the ground. If your AI needed a 3D view of a place by Friday, would you wait for a mapping crew or pay someone nearby with a phone?