Human-robot collision predictor for flexible assembly

Abstract

<p class="Abstract"><span lang="EN-US">The performance of human–robot collaboration can be improved in some assembly tasks when a robot emulates the effective coordination behaviours observed in human teams. However, this close collaboration could cause collisions, resulting in delays in the initial scheduling. Besides the commonly used acoustic or visual signals, vibrations from a mobile device can be used to communicate the intention of a collaborative robot (cobot). In this paper, the communication time of a virtual reality and depth camera-based system is presented in which vibration signals are used to alert the user of a probable collision with a UR5 cobot. Preliminary tests are carried out on human reaction time and network communication time measurements to achieve an initial picture of the collision predictor system’s performance. Experimental tests are also presented in an assembly task with a three-finger gripper that functions as a flexible assembly device.</span></p>