Back to search

University of Waterloo

Advancing Applications in Fluid Powered Artificial Muscle Technology Through Artificial Intelligence Modeling and The Development of a Posture Sensing System

Abstract

dc:description.abstract

This thesis aims to improve human posture by exploring the development and integration of Pneumatic Artificial Muscles (PAMs) and intelligent sensing systems. The objective is to not only develop an efficient posture corrector but also to make a meaningful contribution to the fields of PAM and wearable technology research in the format of delivering a tool to enable the design and optimization of PAMs for wearable applications. To achieve these objectives, three primary projects are designed that show a cohesive progression in the creation and application of these technologies. As compared to other actuators, PAMs can output larger forces which are influenced by many parameters such as their geometries and manufacturing methods. Developing a functional tool for designing and optimizing PAMs is not trivial. The first project involves the creation of ForceSight, by leveraging AI advancements, a tool enabling designers to accurately size PAMs based on specific force requirements. ForceSight predicts the force output of various actuator geometries, thereby simplifying the design process and enhancing the customizability of PAMs for diverse applications. Inertial Measurement Units (IMUs) are sensors used in wearable systems to detect body motion using accelerometers, gyroscopes, and magnetometers. In this thesis, IMUs are used to determine the sagittal slouch and shoulder rounding states of a human subject for posture correction. The focus of the second project is to use a Fuzzy Inference System (FIS) is used to classify the IMU data. This system utilizes data from two 9-axis shoulder-mounted IMUs, emphasizing magnetometer data to assess shoulder rounding, and a lumbar IMU to monitor sagittal slouch posture. The final FIS reliably detects compound slouching motions, providing comprehensive posture assessment based on the sensor data. To validate the AI tool for the design and optimization of PAMs for wearable applications, the third project utilizes the ForceSight tool for determining suitable actuator geometries for a posture corrector and employs the FIS to evaluate shoulder rounding. This section also demonstrates the workflow and benefits of using ForceSight, highlighting its open-source nature and the potential for expanding its data pool to enhance prediction accuracy and applicability across various fields. The posture corrector is an example of evaluating the effectiveness of the developed AI tool, which has broad implications across many industries such as aerospace, industrial automation, and wearable devices. It is the hope of the researchers involved that the technologies demonstrated within this thesis can increase the implementation of PAMs in new areas and use cases across the world.

Degree

thesis:*
Grantor dc:publisher
University of Waterloo
Year dc:date.issued
2024

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Savage, Jordan
Advisors dc:contributor.advisor
  • Ren, Carolyn
  • Dickerson, Clark

Subjects

dc:subject × 12

Rights

Language dc:language.iso
en

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/10012/20800

Chain of custody

source
Harvested from
University of Waterloo
Base URL
dspacemainprd01.lib.uwaterloo.ca/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Savage, Jordan. Advancing Applications in Fluid Powered Artificial Muscle Technology Through Artificial Intelligence Modeling and The Development of a Posture Sensing System. University of Waterloo, 2024. https://hdl.handle.net/10012/20800