Harvesting is the operation where the grape clusters are collected from the vine and are put to containers. While there are mechanical means for grape harvesting, when the maker targets for high quality wine, manual harvesting was the only way… until BACCHUS. Do do this with a robot we made a custom platform with 2 arms.

Harvesting plaform

Robotnik modified an XL platform based on Vogui and place two UR10e arms on top. One arm will carry a customized gripper to grasp the crops, while the other will carry a cutter and a stereo camera. There are other stereo cameras and laser sensors for crop detection and autonomous navigation. Through the cameras the robot has the ability to perform semantic segmentation of the scenes and identify the grape clusters, their stem, leaves and other elements such as poles, branches, irregation tubes etc. All these information are necessary for the robot to navigate and perform autonomous harvesting.

Motion of the robot and its two arms

Imagine the harvesting robot navigating in the middle of the vineyard lane. Using the sensors the platform detects areas of interest where there are mature grapes clusters and approaches them by moving the platform closer and, at the same time, moving the arms close to them.

When the robot comes close to the vine, one arm with the cutter performs visual servoing so that it moves precisely towards the stem of the grape. The other arm approaches the cluster according to the grasp planner and securely grasps it without damaging the grapes.

While the two arms approach to grasp and cut the grape, the entire kinematic chain is controlled in a way that any redundancy is utilized for achieving high dexterity and for avoiding collisions with the environment as well as self-collisions of the robot.

If there are leaves in front of the camera held at the wrist of the cutting arm, this arm moves so that it unveils the partially occluded target. If the leaves obscure the entire view of the camera (e.g. the camera sees the grape but cannot see its stem), the other arm uses an AI method to push them aside and clear the view.

To ensure the high quality of the wine, the harvesting robot will only harvest mature clusters that have been identified and mapped from the inspection platform that carries the hyperspectral camera.

Project Coordinator
Prof. Zoe Doulgeri
Automation & Robotics Lab
Aristotle University of Thessaloniki
Department of Electrical & Computer Engineering
Thessaloniki 54124, Greece

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 871704.

The website reflects only the view of the author(s) and the Commission is not responsible for any use that may be made of the information it contains.

Bacchus Project BACCHUS Project
© 2020 - 2022 All rights reserved