Biologically-Inspired Legged Robot Locomotion Controlled With a BCI by Means of Cognitive Monitoring

Patricia Batres-Mendoza,Andres Espinal,Erick Israel Guerra-Hernandez,Eduardo Perez-Careta,Horacio Rostro-Gonzalez

doi:10.1109/access.2021.3062329

Patricia Batres-Mendoza, Andres Espinal + Show 3 more

Open Access

https://doi.org/10.1109/access.2021.3062329

Copy DOI

Abstract

Brain-computer interfaces (BCI) are a mechanism to record the electrical signals of the brain and translate them into commands to operate an output device like a robotic system. This article presents the development of a real-time locomotion system of a hexapod robot with bio-inspired movement dynamics inspired in the stick insect and tele-operated by cognitive activities of motor imagination. Brain signals are acquired using only four electrodes from a BCI device and sent to computer equipment for processing and classification by the iQSA method based on quaternion algebra. A structure consisting of three main stages are proposed: (1) signal acquisition, (2) data analysis and processing by the iQSA method, and (3) bio-inspired locomotion system using a Spiking Neural Network (SNN) with twelve neurons. An off-line training stage was carried out with data from 120 users to create the necessary decision rules for the iQSA method, obtaining an average performance of 97.72%. Finally, the experiment was implemented in real-time to evaluate the performance of the entire system. The recognition rate to achieve the corresponding gait pattern is greater than 90% for BCI, and the time delay is approximately from 1 to 1.5 seconds. The results show that all the subjects could generate their desired mental activities, and the robotic system could replicate the gait pattern in line with a slight delay.

Highlights

Brain-Computer Interfaces (BCIs) are systems that provide a communication and control channel between the human brain and the outside world by means of electroencephalography (EEG) [1]
Our work aims to control a bio-inspired hexapod robot in real-time with the help of a BCI based on motor imagery
It includes the Motor Imagination (MI) training process to familiarize the user with the imagined commands using visual stimulation, receive feedback to speed up the learning, and improve the real-time experiment’s performance

Summary

Introduction

Brain-Computer Interfaces (BCIs) are systems that provide a communication and control channel between the human brain and the outside world by means of electroencephalography (EEG) [1]. BCIs are developed to help patients suffering from severe motor impairments [2]–[4]. Several applications have emerged outside of the medical field, like the integration of BCIs with other immersive technologies such as virtual reality (VR), augmented reality (AR), and computer games [5]–[8]. There are researches related to integrating BCIs and external devices to robot systems [9]. In this last case, researchers have used several strategies to control a robot with a BCI such as Visual Evoked. Potentials (VEPs) [10]–[13], Event-Related Potential (ERP) [14]–[18], Slow Cortical Potentials (SCPs) [19]–[21] and Sensorimotor Rhythm μ and β [22]–[24]

Methods

Results

Conclusion