Trajectory optimization applied to motion planning of industrial manipulators.

Abstract

Robotic manipulators are becoming more present in various activities, inside and outside the industry. The use of these robots allows greater precision and accuracy in carrying out the tasks, however, it is important to take into account some factors to ensure the safety of the system, such as the ability to avoid obstacles that may be present in the operating environment and the quality of the final path. In this work, we implement a system for trajectory optimization of a robotic manipulator in static environments using the algorithms Covariant Hamiltonian Optimization for Motion Planning (CHOMP) and Stochastic Trajectory Optimization for Motion Planning (STOMP). We integrated an RGB+D sensor to the system for obstacle detection on manipulator's workspace. The system is based on open-source framework Robot Operating System (ROS) and it is applied to pick and place tasks in an additive manufacturing cell composed by the collaborative robot UR5. After a series of executions on real and simulated scenarios, the algorithms were compared based on their success rate, planning time, and duration of the generated trajectory. Results indicate that the proposed system can generate feasible and collision-free trajectories in static environments.