Finite element optimization of a flexible fin-ray-based soft robotic gripper for scalable fruit harvesting and manipulation

Abstract

On the path to achieving fully autonomous farming, the use of grasping devices for fruit picking and handling remains an open challenge. Current solutions are designed for specific fruits and robot manipulators, often without considering the intrinsic interaction between the gripper's fingers and the fruit. This work explores the use of fin-ray-based flexible grippers, which mimic human fruit-picking movements, for harvesting and pick-and-place operations involving medium-sized fruits. Optimal gripper characteristics were determined through a Finite Element Analysis methodology. To achieve the harvesting objective, the grippers were integrated into a vision-based system and a robotic manipulator, with testing conducted under laboratory conditions. The harvesting study focused on apples, while the manipulation task was tested with apples, oranges, and lemons. The findings indicate that while all grippers demonstrated a suitable performance, one particular design emerged as the most effective, meeting all criteria and outperforming the others in experiments and performance metrics.