Dragonfly is a computer vision based technology and allows to simultaneously localize robots, AGVs, drones, forklifts (and any other asset that can be equipped with just a monocular or stereoscopic camera) while improving the localization accuracy through a continuous mapping of the surrounding environment (SLAM – Simoultaneous Localization And Mapping).

Dragonfly estimates in real-time the 3D position of the camera and provides its latitude, longitude, altitude and orientation (6 Degrees of Freedom).

Dragonfly represents the state-of-the-art in indoor localization technologies (where systems such GPS cannot be used). Compared to Dragonfly, other indoor localization technologies are:

• less competitive in terms of installation and maintenance costs (laser systems, UWB).
• less competitive in terms of localization accuracy (systems based on the measurement of the attenuation of WiFi/Bluetooth radio signals).

How it works

While Dragonfly is running:

• a camera captures real-time images of the environment and sends them to a computing unit.
• the geo-localization algorithm inside the computing unit takes care of:
  1. extrapolate the peculiarities of the environment detected in each individual frame.
  2. compare the real-time detected peculiarities with the peculiarities saved inside a 3D map of the environment (created previously in the mapping phase).
  3. compute the position of the camera in the form of WGS-84 coordinates (latitude and longitude) and distances in meters (X, Y) with respect to an origin and make them available through
     • a web user interface (accessible via a browser).
     • a REST API.