Robotic Green Fruitlet Thinning with a Soft Manipulator (Arm)-based Robotic System

Moonshot Award

Hybrid Award (Proposal Development & Innovation Development)

Abstract

To achieve desired fruit size and quality (e.g., color, sugar content), excess young fruit (fruitlet) is removed from apple trees in the early growing season. Currently, this green fruitlet thinning operation is performed almost entirely by hand, making it one of the most labor-intensive and costly tasks in orchard production. Our moonshot project is developing an intelligent robotic system that combines machine vision, a gentle soft robotic arm, and a precision cutting tool to identify and selectively remove targeted fruitlets with minimal damage to the tree. We have successfully designed, built, and integrated a working indoor prototype that can locate fruitlets, navigate toward their stems, and perform selective removal. 

The next milestone is to validate the system in commercial orchards, where natural conditions such as wind, changing sunlight, dense foliage, and hidden fruit make autonomous operation much more challenging. Because green fruitlet thinning is only possible during a brief biological window of about three weeks each spring, field testing must occur during this limited period. Successful completion of this phase will move the technology closer to practical orchard deployment.