Collision free path planning of a robotic manipulator for pruning apple trees

Pruning of apple trees requires 80-120 working hours of labor per hectare accounting for 20% of the total production cost. Robotic pruning is a potential solution to decrease labor dependence and associated costs. Autonomous precise manipulation of a robotic manipulator in presence of obstacles is a challenge. The spatial requirements and collision-free path planning for the robotic manipulator is essential for automated systems. This simulation study focused on investigating the branch accessibility of a six-rotational (6R) degrees of freedom (DoF) robotic manipulator with a shear blade type end-effector. A virtual tree canopy environment was established in MATLAB for simulation. The Rapidly-exploring Random Tree (RRT) obstacle avoidance algorithm was used to establish a collision-free path to reach the target pruning points. The path smoothing and optimization algorithms were also used to reduce path length and calculate the optimize path. The simulation showed that the integrated robotic manipulator reached the pruning points avoiding obstacle untargeted branches. The path generation time, path length, target reaching time, and number of accessible branches (success) and collisions (failure) was recorded. The study provides the foundation information for future work on the development of a robotic pruning system.