Workspace-Based Approach for Robot Trajectory Evaluation

A robot’s workspace refers to the surrounding environment where robot can move, operate, and interact with the other objects for task accomplishment. Analysis of robot workspace is an important aspect in robot’s motion planning, task execution, and interaction with the world. This paper presents a framework to assess the suitability of a given trajectory in the robot workspace. The framework utilizes kinematic modeling along with the adaptive neuro fuzzy inference system to learn the robot workspace model. Afterward, the learned model is used to retrieve an analogous trajectory corresponding to the specified task at the trajectory level. The approach is demonstrated for a six degrees of freedom industrial robot in a simulation environment. Simulation results reflect that the proposed approach provides a simplified way to evaluate the executability of an externally defined trajectory by the robot.