Working prototype
COFRO
Semi-automated coffee harvesting
A handheld device that picks coffee cherries and separates ripe from unripe as it goes, so the sorting happens in the field rather than afterwards.
An object-detection model trained across seven classes of coffee cherry decides what to pick and which bin it belongs in. It runs on a Raspberry Pi 4 with a Hailo-8L accelerator, because the decision has to be made in the time it takes to close a gripper — not after a round trip to a server that a coffee estate may not be able to reach.
- Object detection
- Edge inference
- Raspberry Pi 4
- Hailo-8L

COFRO




