A Universal Self-Adaption Workspace Mapping Method for Human–Robot Interaction Using Kinect Sensor Data

Abstract

Human-robot interactions (HRIs) are a key technology to combine human and robot capabilities. Though many feasible methods have been proposed in this field, the current workspace mapping methods are deficient in terms of generality. In this paper, a universal control system based on the self-adaption workspace mapping method for HRI is designed. In normal circumstances, human movement is adjusted adaptively in a safe, obedient, stable and efficient way to the external environment. The universal control system could naturally carry out an object manipulation task by allowing a human operator to communicate his motions to the robot manipulators. More importantly, by applying this system, people no longer need to find parameters by trial-and-error because they are not affected by changes in the human and robot model. The function of similarity that is used to evaluate the similarity between a human arm and the robot is introduced. Moreover, experiments are carried out, and an analysis of results verifies the feasibility and excellent performance of the universal control system. Finally, a demonstration is applied to display the flexible implementation of the object manipulation task that is collaboratively done by the universal robot through this system, and the method is evaluated by comparing it with other method.