The spatial foundation
for Physical AI
DeepHub® harmonizes functional areas, assets, real-time position and object orientation from any positioning source, and exposes them as agent-ready tools through the Model Context Protocol (MCP). This is the spatial ground truth every Physical AI agent needs, and no team should have to build it on their own.
Vendor-neutral and model-agnostic. Indoor and outdoor. Built for end-to-end value chains, from the warehouse and the factory floor to the supply chain and the production line, down to the single asset movement.
We co-initiated omlox, the open locating standard, and have built location infrastructure for industry for more than twenty years. The DeepHub® MCP is that groundwork, made callable by an agent.
What the layer exposes

One interface, irrespective of which positioning technology produced the data

Functional areas and geofences, defined and queried through standardized MCP

Assets and their identity across systems

Position and object orientation in real time

Dwell time per functional area

Proximity between assets

Reading and writing, for example creating a fence or registering a trackable
We not only crafted the layer. We built our own agent on top of it.
Interact with your facility
The DeepHub Agent is our own agent, built entirely on the same layer. Point it at your DeepHub, ask it questions the way you would ask a colleague who never leaves the floor, and let it make the changes that follow from the answer.
Built on the layer we ship
The DeepHub Agent has no private access. It runs on the same MCP tools every team building on the spatial foundation receives, the same calls that define and query functional areas, read asset position and object orientation in real time, compute dwell time and proximity, and write changes back.
The core concept behind DeepHub has always been technology interoperability. With the MCP interface, we ensure maximum flexibility even when used with agents.

DeepHub® Agents
Enables you to deploy autonomous agents based on insights gained through DeepHub® Analytics.





