Physical AI and 3D capture data.

Real spaces as data for machines.

Physical AI — robotics, autonomous systems, and simulation — increasingly trains and tests against digital reconstructions of real environments. A robot learns to navigate a warehouse, a planner simulates a facility, a model is validated against a faithful copy of a real space. All of it needs spatial data that reflects the world accurately.

Handheld LiDAR is one practical way to capture that data. A walk through a building produces a measurable point cloud; the same space can become a mesh or a photorealistic 3DGS scene — exactly the kind of reference simulation and training pipelines consume.

Vendors like NVIDIA build tooling in this space. Volumet is not affiliated with them. The relevant point is upstream: before any of that tooling runs, someone has to capture the real environment — and that is where a scanner rental fits.