Intelligent Interactive Control via Haptic E‐skin for Human–Robot Interaction and Collaboration

Guozhen Li,Rong Zhu

DOI: https://doi.org/10.1002/aisy.202200101

IF: 7.298

2022-10-07

Advanced Intelligent Systems

Abstract:An intelligent interactive control method is proposed based on generalized virtual spring‐damping force‐touch mechanism using e‐skin, which converts contact force and contact position to generalized forces for manipulating robot multidimensional movements. Human–robot‐in‐the‐loop architecture is utilized to accomplish robot task learning through hand‐by‐hand teaching wherein human teaches robot to pick, carry, and place objects in an easy and flexible way. Human–robot interaction and collaboration are emerging as critical and necessary technologies as robots become more commonplace in our lives. Haptic perception and intelligent control are essential to ensure safe and dexterous interactions between humans and robots. Traditional interaction methods using robot proprioceptors rely on robot dynamic models that complicate robotic controls and narrow their adaptabilities. Herein, an intelligent interactive control method is proposed based on a generalized virtual spring‐damping force‐touch mechanism using haptic e‐skin, which converts contact forces and contact positions on the e‐skin to generalized forces for implementing robot compliance controls of multidimensional movements. By touching the e‐skin mounted on a robot arm, a human can readily manipulate the robot arm, achieving flexible movement and rotation. An intelligent architecture of human–robot‐in‐the‐loop is proposed to accomplish versatile interactive and collaborative actions. A task learning is demonstrated for a robot from human, that is, a human teaches a robot how to pick up, carry, and place an object into a box by hand‐by‐hand teaching, which shows promising prospects in intelligent robot applications.