Technology · Architecture

How FRP fits together.

Four core components, one protocol between robot and cloud, and a clean separation between the people who manage access and the robots that do the work.

FRP architecture: robot software connects outbound over SRCP to the broker in the FRP cloud; the broker connects to the FRP backend, which serves the FRP frontend to teams in any browser; external apps connect to the broker with scoped credentials; the Digital Twin runs alongside the robot software and Core-Edge Inheritance connects to the broker

On the robot

Robot software

A compact C++ client on the robot’s computer (Raspberry Pi 5 or NVIDIA Jetson class). It connects outbound to the broker and runs the robot.

In the cloud

Broker

The SRCP hub. Robots, the platform and authorised external apps all connect to it; it routes telemetry, video, control and maintenance traffic.

In the cloud

FRP backend

Users, teams, keys, roles, projects and data. It decides who may talk to which robot.

In your browser

FRP frontend

The web application where teams develop, operate and train robots.

Alongside the robot

Digital Twin

In development. The robot’s own software connected to a virtual 3D world.

Coming

Core-Edge Inheritance

Our technology for managing AI across the fleet: a central AI that passes its knowledge to compact models on each robot.

Open to your software

Build your own applications on FRP.

External applications, such as a high-performance native control station, connect directly to the broker and talk to robots over SRCP. Access is managed centrally by the FRP backend and scoped to specific users and robot groups.

  • Direct broker connection for performance
  • Scoped credentials, never the master key
  • Same protocol and security model as the platform itself

See FRP on your robots.

Tell us about your robots, your classroom or your mission. We will show you the platform end to end.