Transparency-inspired tools for wearable robots

• Ano de defesa: 2021
• Autor(a) principal: Noppeney, Victor Tamassia
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: compatibilidade cinemática; controle de transparência; exoskeleton robots; kinematic compatibility; motion prediction; predição de movimentos; robôs exoesqueleto; robótica; robotics; transparency control
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/18/18162/tde-23072024-084027/

Exoskeleton robots are characterized by their anthropomorphism and operation in intrinsic contact with a human being, i.e., the user. Most of their applications include expanding the physical limits of human beings, whether in the rehabilitation of people with locomotion limitations, in the reduction of metabolic rate during the execution of a given task, in the carrying of loads with high mass, or in the amplification of the user\'s force. However, such applications are only possible if the exoskeleton moves in harmony and synchrony with its user, so that the user does not feel discomfort nor need to expend additional effort while using the wearable robot. This concept is called transparency, and is achievable through the combination of sensing, estimation, control, and actuation systems. This work will study the proposals that seek such transparency in the State of the Art and propose new ways to pursue this goal, focusing mainly on kinematically compatible mechanical design and human motion prediction for anticipatory behavior. Regarding mechanical design, different concepts focused on biomimetics and ergonomics are discussed and put into practice through the design of ankle and knee joints. After that, a universal human motion prediction algorithm, based on artificial neural networks, is proposed and evaluated.