Overview of control systems for robotic harvesting of sweet peppers and apples

Abstract

Automated harvesting systems are becoming a crucial part of digital agriculture. Substituting time-consuming and laborious manual harvesting with a continuously automated operation would result in reduced human efforts, which contribute to higher field efficiency. This could be achieved by means of robotic harvesting which comprises robot manipulators, gripping and grasping mechanisms, software implementations. However, the inadequate design of control strategies could cause the agricultural production loss. This paper reviews some of the latest achievements in control systems of agricultural robotics and more specifically in robotic harvesting. The employed robot arms, their degrees of freedom (DOF), and the crops are also considered in this review study. While the control algorithms are being developed with high robustness and fast-response properties, our conclusion is that the performance of controllers could be drastically affected by different parameters such as the number of DOF, estimation accuracy of the robot pose during visual servoing and failure of robot’s inverse kinematics solver.