Augmented online point to point trajectory planning, a new approach in catching a moving object by a manipulator

In a structured dynamic environment, due to the known variation of the environment, robot trajectory planning can be performed robots motion starts. Within an unstructured dynamic environment, environment variations are permanent, so that the surroundings information must be acquired in an online action. Likewise the data acquisition, robot trajectory planning must be executed online. In this paper a method for online trajectory planning based on adaptive prediction planning and execution (APPE) is proposed for robotic manipulators considering velocity and torque constraints. A vision-based data acquisition system is used to acquire object positions and velocities. To describe the end-effector trajectories in joint and Cartesian space a set of time polynomial function are used. Initially a primary trajectory is planned and it will be updated in definite time periods. This action will be frequently repeated until catching is occurred. Proposed method has been tested on an experimental setup and the experimental results are presented.