Tutorial – Half Day (Great Room 3 – Capacity: 80)

Title:

Biosignals-based design approaches for the development of human-machine interfaces for shared control of computer applications and robotic devices

Abstract

Various human machine interfaces (HMI) have been developed employing various control methods to suit
specific needs of the user. These interface can employ various types of biosignals to offer an intuitive and
hands-free method of robotic control. However, biosignals require pre-processing and can vary
depending on biological differences of the users leading differences between individuals as well as the same individual between recording sessions. The content covered in this proposed tutorial is based on a very recent work of the authors where they developed semiautonomous and shared control schemes based on biological signals from the user. Additionally, the tutorial also covers the data collection procedure and data treatment steps in order to train machine learning models to decode these signals. Therefore, the tutorial will provide the participants with all the tools needed to perform data collection, model training and development, and implement control of HMI to perform a semiautonomous and shared control for execution of complex manipulation tasks using robotic devices.

Agenda

Format: Half day tutorial
Time: August 26, 2023

The tutorial agenda will be as follows:

• Background of Biosignals, 1:00 pm – 1:40 pm
Speaker: Anany Dwivedi
A presentation focusing on the working principles of biosignals such as electromyography, forcemyography, and lightmyography. In sequence, there will be a comprehensive description of data collection and signal pre-processing methods common in the field.

• Machine Learning Models for Biosignals, 1:45 pm – 2:30 pm
Speaker: Ricardo Vilela De Godoy
A presentation session focusing on how to develop, train, and deploy some classic machine learning
techniques, as well as novel deep learning methods, from data collection to model evaluation.

• Break, 2:30 pm – 3:00 pm

• Gesture Decoding Demonstration, 3:00 pm – 3:45 pm
A demonstration of the control of a robotic hand using electromyography.

• Developing Shared Control Interfaces with Different Biosignals, 3:45 pm – 4:15 pm
Speaker: Bonnie Guan
A presentation session focusing on the control of robotic devices for the development of shared control
and semiautonomous frameworks using different biosignals. This session will also cover different applications, such as telemanipulation tasks with robot arm-hand systems, control of prosthetic devices,
etc.

• Open Discussion, 4:15 pm – 5:00 pm
Open floor to discuss biosignal-based human machine interfaces

Organizers

Bonnie Guan, Doctoral Candidate
New Dexterity Research Group, Department of Mechanical and Mechatronics Engineering, The University of Auckland, Auckland, New Zealand
E-mail: bgua324@aucklanduni.ac.nz
Phone: +64-223097768

Ricardo Vilela De Godoy, Doctoral Candidate
New Dexterity Research Group, Department of Mechanical and Mechatronics Engineering, The University of Auckland, Auckland, New Zealand
E-mail: rdeg264@aucklanduni.ac.nz
Phone: +64 029 020 78917

Anany Dwivedi, Postdoctoral Fellow
Artificial Intelligence (AI) Institute, Division of Health, Engineering, Computing and Science, The University of Waikato
E-mail: anany.dwivedi@waikato.ac.nz
Phone: +64-210411362

Minas Liarokapis, Associate Professor
New Dexterity Research Group, Department of Mechanical and Mechatronics Engineering, The University of Auckland, Auckland, New Zealand
E-mail: minas.liarokapis@auckland.ac.nz
Phone: +64 93737599 Ext.86688