Bloch, V., Division of Environmental Water and Agricultural Engineering, Faculty of Civil and Environmental Engineering, Technion, Haifa, Israel, Agricultural Research Organization (ARO), Volcani Center Agricultural Engineering, Bet Dagan, Israel
Degani, A., Division of Environmental Water and Agricultural Engineering, Faculty of Civil and Environmental Engineering, Technion, Haifa, Israel
Bechar, A., Agricultural Research Organization (ARO), Volcani Center Agricultural Engineering, Bet Dagan, Israel

Manipulation tasks in precision agriculture require a robotic arm that can perform the task quickly and precisely. Robots with kinematics suitable for a specific task will perform better than a universal robot expected to perform several tasks. To provide a good fit between a robot and its task, the robot design must be based on a well-defined task model. A unique approach is proposed to design an optimal robot for a specific precision agriculture task and an approach to the classification of such tasks. One of the tasks considered is fruit picking. Apple and nectarine trees were measured with a passive manipulator device developed for this purpose. Statistical analysis was performed to define differences and similarities between the tree types. The results showed that different tasks require different robots to perform optimally in each task. In addition, precision agriculture tasks can be classified according to the optimal robotic arms.