An Intuitive Teleoperation of Industrial Robots: Approach Manipulators by Using Visual Tracking Over a Distributed System

An Intuitive Teleoperation of Industrial Robots: Approach Manipulators by Using Visual Tracking Over a Distributed System

Andrea Bisson, Stefano Michieletto, Valentina Ferrara, Fabrizio Romanelli, Emanuele Menegatti
Copyright: © 2019 |Pages: 19
DOI: 10.4018/978-1-5225-8060-7.ch050
OnDemand:
(Individual Chapters)
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

Teleoperation of manipulator robots with RGB-D sensors is now mainly done using inverse kinematics techniques. In this chapter, we describe an intuitive way to teleoperate an industrial manipulator through vision sensors by directly controlling manipulator joints retargeting specific human motion. In this way the human operator has the full control of robot movements with practically no training, because of the intuitivity of this teleoperation method. The remapping into the robot joints is done by computing angles between vectors built from positions of human joints, tracked by the selected vision sensor. The obtained system is very modular which allows to change either the tracking sensor or the robot model with some small changes. Finally, the developed teleoperation system has been successfully tested on two real Comau robots, revealing to be fast and strongly reliable.
Chapter Preview
Top

The Developed Teleoperation System

In this Section we will describe how the developed teleoperation system works. First of all we will describe how the physical system used is composed and interconnected, then we will focus on how the developed algorithm works.

The System Architecture

The developed system architecture is composed by two computers:

  • A common Ubuntu Linux computer connected to the Microsoft Kinect sensor, on which is installed ROS middleware.

  • A real-time Linux computer connected both to the computer previously described and also to the robot controller.

Complete Chapter List

Search this Book:
Reset