Not all fruits are visible to a fruit picking robot

Not all fruits are visible to a fruit picking robot

Selective harvesting robots now mainly use camera sensors to determine where, for example, an apple or pear hangs. But cameras don't see everything: the fruit could be hidden behind a leaf or branch, for instance. In that case, other techniques are required, such as sensors that measure pressure. The Greenhouse Horticulture and Flower Bulbs Business Unit of Wageningen University & Research and Oregon State University are investigating how information from various sensors can be merged and interpreted, and how this can be used to control the gripper of a robot.

The project 'Integrated sensing and real-time control for intelligent fruit picking' is funded by the American Washington Tree Fruit Research Commission (WTFRC), and is also a project of the Agro Food Robotics programme. In the 3-year project, the integration of data from various sensors, i.e. vision, force and tactile sensors, is sought. For example, it is possible that a robot – just like a human does – feels whether it is grasping the fruit in the right way. This information must be translated into the correct control of the robot arm and the gripper by an algorithm, so the robot can harvest the fruit. Both techniques (software and hardware) will be tested at the WUR experimental orchard location in Randwijk.

Continue reading.

Photo: WUR

Source:

Share