Optimal conceptual design and vision-based control of a fruit harvesting robot

The main contribution of this paper is to develop a vision-based control of a robotic arm for the harvesting fruits. The camera fixed in the gripper pad enables to precisely locate the fruit and pluck it from the branch. Rigorous stability analysis is done to ensure the guaranteed performance of the closed loop system. The camera feedback locates the exact position of the fruit; this enables the controller to track a suitable and optimal path to reach the target by performing desirable transformations. The manipulator with five-DOF (RRPRR) is designed and optimised for the formulating simple control strategies. The finger-like built gripper is electrically actuated to provide the necessary force required in harvesting the fruit. Also an additional bellow kind of structure is specially designed and located below the gripper which helps to roll down the harvested fruit on to the storage container without damaging it. Numerical simulation analysis was carried out along with the design realisation to justify the context. The advancement in the field of agrionics has also been a source of inspiration in designing agricultural robots.