A robot, a headset, and a digital twin can all hold representations of the same door—and still have no shared way to say, “this one.” Once that door is grounded, 4D-ID gives it a persistent identity they can all resolve.
A map calls it a feature. A scan sees a cluster of points. A building model stores an asset. Each system has a useful view. None of those views, on its own, tells the others: we mean the same thing.
That gap turns every handoff into another matching problem. Before systems can share what they know, they have to work out what they’re talking about.
Coordinates tell you where. Identity tells you which.
A position can tell a robot where to go. It cannot, by itself, tell the robot that the door in today’s scan is the door inspected last month.
4D-ID separates those questions. Give the thing a persistent name. Let each system keep its own coordinates, representations, and tools. Use the shared name to connect them.
Keep your maps. Keep your models. Share the name.
Less matching. More doing.
The payoff isn’t another identifier in a database. It’s a reference that survives the handoff:
- A drone submits damage evidence. After grounding, its observation refers to the named fixture—not just a point in a video.
- A robot receives the job. The repair request points to that fixture, even though the robot navigates with a different map.
- An AI checks the history. It can retrieve observations and repairs attached to the same identity, rather than treating each sighting as a new thing.
A name does not align sensors, prove an observation, or plan a safe route. It gives those systems a stable subject to work on.
Share the identity. Control the information.
Agreeing on which door does not mean sharing everything known about it.
A public map, a maintenance team, and a security system can refer to the same identity while holding different information. The identity is shared. Access to records, precision, and private layers is controlled separately.
One name does not mean one database—or one view of the world.
What if two systems name the same thing?
They may. Systems can generate candidate identifiers independently, so agreement cannot depend on everyone starting in the same place.
Start with identifiers already in use: an asset tag, a parcel number, or a building-model identifier. Spatial matching can suggest candidates. Further evidence can establish whether two records describe the same thing, so duplicate names can be merged.
That still takes evidence. Two nearby objects are not necessarily the same object. Shared identity gives matching a durable result; it does not make matching automatic or infallible.
A shared layer, not a replacement stack.
4D-ID builds on the standards that already describe space: global grids for indexing, GeoPose for position and orientation, and existing tools for maps, models, and movement.
The aim is not to make every system describe the world the same way. It is to let them refer to the same thing.
One shared name. Many independent views. A foundation for systems that can work together.
From grounded observation to action, by identity.
Watch a supplied drone observation attach to a known fixture and pass to a robot. The scene simulates localization and navigation separately.