Brain-controlled robot agent: An EEG-based eRobot agent

Li Wei Wu*, Hsien Cheng Liao, Jwu-Sheng Hu, Pei-Chen Lo

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

4 Scopus citations

Abstract

Purpose - This paper aims to present a novel embedded-internet robot system based on an internet robot agent and the brain-computer interface (BCI) scheme. Design/methodology/approach - A highly flexible and well-integrated embedded ethernet robot (eRobot) was designed with enhanced mobility. In the eRobot, a circuit core module called a tiny network bridge (TNB) is designed to reduce robotic system cost and increase its mobility and developmental flexibility. The TNB enables users to control eRobot motion via embedded ethernet technology. Through electroencephalogram (EEG) feedback training, the command translation unit (CTU) and alertness level detection unit (ADU) allow the eRobot to perform specific motions (for example, lying down or standing up) to reflect alertness levels of the user, and move forward, turn left or right following the user's command. Findings - After a short training period, subjects could achieve at least 70 percent accuracy in the CTU game testing. And the error rate of ADU, estimated from the results of classifying 496 labeled EEG epochs, was approximately 10.7 percent. Combining an encoding procedure, the commands issued from the CTU could prevent the robot from performing undesired actions. Originality/value - The eRobot could reflect some physiological human states and be controlled by users with our economical design and only two bipolar EEG channels adopted. Thus, users could make the EEG-based eRobot agent his or her representative. Based on the proposed EEG-based eRobot system, a robot with increased sophistication will be developed in the future for use by disabled patients.

Original languageEnglish
Pages (from-to)507-519
Number of pages13
JournalIndustrial Robot
Volume35
Issue number6
DOIs
StatePublished - 10 Nov 2008

Keywords

  • Computer networks
  • Control systems
  • Man machine interface
  • Robotics

Fingerprint Dive into the research topics of 'Brain-controlled robot agent: An EEG-based eRobot agent'. Together they form a unique fingerprint.

Cite this