Farming Autonomy Operating System

Hybrid Localization

Always knows where it is, with or without GNSS

Your machine always knows where it is. Hybrid localization fuses LiDAR and IMUs with GNSS and RTK when they are available, and keeps working wherever the signal gets too poor to use.

How it works

  1. 01

    RTK GNSS where the sky is open

    Out in the open field, RTK GNSS gives your machine its position to the centimetre, and the FAOS uses it.

    Lero.03 drives along a vegetable bed in the open field towards a farm warehouse. Satellites hang in the sky, each with a dotted signal running down to the antenna on Lero's roof, and its position fixes trail behind it in a tight line: an RTK fix, to the centimetre.
  2. 02

    Where GNSS alone gives up

    Near buildings, under tree canopies or inside a barn or warehouse, the satellite signals are weakened and bounce off walls and roofs, and the position GNSS gives drifts by metres. A machine that relies on GNSS alone can no longer trust it and has to stop.

    Relying on GNSS alone, Lero.03 drives on into the warehouse's open door. One satellite's signal after another has to pass through its walls and roof, marked with a warning where it meets them, and reaches the antenna only faintly. Its position fixes scatter by metres, and the ring showing how unsure it is grows, turns orange and drifts off the robot, until it stops in the door.
  3. 03

    LiDAR alongside GNSS

    The FAOS scans the ground and the buildings around your machine with LiDAR as it drives. As the satellite signals degrade, it trusts GNSS less and leans on its scans, so it stays sure of where it is.

    The stopped robot shrinks into a card as the same drive starts again with the FAOS. Rings of LiDAR returns spread out around Lero.03 and its scans pick out the warehouse's walls ahead. As it drives in and the satellites' signals weaken at the building one after another, the estimate of where it is stays tight around the robot.
  4. 04

    Drives on when GNSS can't be used

    Once GNSS is too poor to use, the FAOS matches its LiDAR scans to the map of the site, its walls and racks, and your machine drives on through.

    Inside the warehouse, past pallet racks loaded with boxes, Lero.03 drives on while every satellite's signal is weakened by the roof and its position fixes scatter around it. The LiDAR returns off the walls and racks turn green as they're matched to the map, and the building's steel frame and racks light up with them.
  5. 05

    Back to RTK without a jump

    Out in the open again, the FAOS picks RTK back up and blends it in smoothly, so the position never jumps.

    Lero.03 drives out of the far end of the warehouse into the open field. The satellites' signals reach its antenna cleanly again and its position fixes tighten back into a line, while the estimate of where it is carries on smoothly, without a jump.