Dragonfly can be split into 3 components:

• a camera (USB, built-in, network connected) that sends the video stream to the computing unit that runs the Dragonfly Application.
• a computing unit (on which it runs the Dragonfly Application) that receives the video stream from the camera and computes the camera’s position.
• a web interface (Dragonfly Web User Interface) that is accessed with a browser and that is used to configure and monitor the Dragonfly Application.

The 3 components are completely separated. This makes Dragonfly highly versatile and lets you choose between different architectures all described in the next sections.

On board local computation

Built-in/USB camera directly connected to the computing unit that runs the Dragonfly App.

• PRO:
  • there is no need to have a reliable Wi-Fi LAN.
  • minimum latency for what concerns the computation of the locations.
  • more reliable because each computation unit is independent and there is not SPOF (single points of failure) for multiple assets.

Off-board local computation

Network camera that makes available its video stream using a Wi-Fi local network to the computing unit that runs the Dragonfly App.

• PRO:
  • the maintenance of the computing units is easier because they are NOT located on board of the assets.
  • the weight on the asset is lower.
  • the power consumption on the asset is lower.