GREEN FRUIT REMOVAL DYNAMICS AND ROBOTIC GREEN FRUIT THINNING SYSTEM

Manual green fruit thinning is a labor-intensive task, and therfore is not practical efficient in a large-scale application. In addition, chemical fruit thinning is not only climate-dependent, but also time-sensitive and cultivar-dependent. Non-selective mechanical thinning may remove good fruits and damage fruits or tree canopies. This project proposes to develop an automated selective fruit thinning system. The main idea is to develop a novel robotic green fruit thinning system that will precisely detect and locate green fruit and selectively remove those unwanted fruits without damaging the remaining fruit. During this project, fruit thinning criteria and fruit removal dynamics will be studied. In order to accomplish the proposed tasks, a machine vision system will be developed to detect the green fruits in a tree canopy and locate fruit cluster regions as well as fruit distribution densities. The collected information from the machine vision system will be used to determine fruit to be removed. Two robotic fruit removing end-effectors will be developed to remove the targeted green fruits selectively. The expected outcomes will be a practical robotic green fruit thinning system that can effectively thin apple trees with selective fruit removal. The implementation of such a system will significantly increase productivity and improve the long-term economic and social sustainability of the U.S. tree fruit industry. Even though the focus of this project will be on apple, the knowledge and technology developed can be potentially expanded to thin other tree fruit crops, such as pears and peaches.