Design, Integration and Field Experiment Evaluation of the Citrus Fruit Harvesting Robot

The research and development of citrus fruit harvesting robots to realize the automation and intelligence of harvesting are of great significance for reducing the labor cost and labor risk of harvesting. In this paper, the citrus fruit harvesting robot with running module, vision module, manipulator module and end-effector module was developed according to the growth characteristics and harvesting coordinate system of citrus trees in an unstructured environment. The robot was hand-eye calibrated, and the Robot Operating System (ROS) was used as the control system to control the citrus fruit harvesting robot. An inverse kinematics selection method for the harvesting manipulator with minimum energy consumption was proposed, based on the method, a function for judging the order of harvesting multiple citrus fruits when the robot was harvesting multiple citrus fruits was proposed and the harvesting process of the robot was determined during harvesting multiple citrus fruits. The superiority of the harvesting sequence calculated by the principle of "minimum energy consumption" was verified in a laboratory environment. The single fruit and multiple fruits harvesting experiments were carried out in the field environment, whose results were evaluated and analyzed. The field experimental results showed that the success rate of the citrus fruit harvesting robot developed in this paper for harvesting single fruit was 87.5% and that for harvesting multiple fruits was 75.71%, and the overall success rate was 80%. To develop more powerful path planning algorithms to improve the robot's ability of perceiving the environment and avoiding obstacles will be done in future