Skip to main content

Lesson

Architecture

How Cyberwave services, runtimes, and data planes fit together.

Cyberwave follows a classic Cloud-Edge topology, optimized for the constraints of physical automation (intermittent connectivity, bandwidth limits, real-time safety).

System Map

  1. Control Plane (Cloud)

    • Backend: A Django-based monolith providing the REST API (django-ninja), Identity Management, and Asset Registry.
    • Orchestrator: Manages mission state, fleet configuration, and OTA updates.
    • Data Lake: Stores historical telemetry and mission logs (PostgreSQL/TimescaleDB).
  2. Data Plane (Transport)

    • MQTT Broker: The nervous system. Handles lightweight, high-frequency telemetry (e.g., /robot/joint_states) and command signals.
    • WebRTC Signaling: Negotiates peer-to-peer video streams between the Edge and the Client Dashboard, bypassing the cloud for lowest latency.
  3. Edge Plane (Device)

    • Agent: A lightweight process (Python/C++) running on the robot's computer (e.g., NVIDIA Jetson, Raspberry Pi).
    • Bridges: Adapters that translate Cyberwave commands into local driver calls (e.g., ROS2 Bridge, MavLink).
  4. Experience Layer (Client)

    • Frontend: A Next.js application delivering the 3D dashboard.
    • Visualizer: A WebGL engine (three.js) that renders the Digital Twin state in real-time.

For detailed deployment topologies, including air-gapped options, refer to the architecture reference.

More tracks

Keep exploring Cyberwave Learn

Every track is defined in markdown so adding new lessons stays fast and versioned.