Abstract

This study introduces a self-powered smart insole system designed for real-time monitoring of foot health, with a specific focus on detecting flatfoot conditions. The insole integrates multiple triboelectric energy harvesters strategically positioned to capture electrical signals generated from ground reaction forces during daily activities such as walking, jogging, and running. Proof-of-concept testing was conducted on a single participant under two conditions: a healthy foot and a simulated flatfoot created by reducing the medial arch height by approximately 70\. In the healthy foot trials, the system demonstrated consistent and reliable performance, with negligible electrical output from the medial arch sensor, as expected due to minimal ground contact in this region. In contrast, the simulated flatfoot condition produced a significant increase in voltage output from the medial arch sensor, successfully identifying the abnormal foot mechanics associated with arch collapse. Additionally, a neural network was implemented to classify healthy and flatfoot conditions from the collected data, achieving an accuracy of 86% and a precision of 96%, demonstrating the feasibility of machine learning integration for automated flatfoot detection. Overall, the findings validate the smart insole's capability as a promising tool for continuous foot health monitoring, early diagnosis of flatfoot, and future applications in personalized rehabilitation and preventative care

Date of publication

2025

Document Type

Thesis

Language

english

Persistent identifier

http://hdl.handle.net/10950/4842

Committee members

Dr.Nelson Fumo, Dr.Chung Goh, Dr.Alwathiqbellah Ibrahim

Degree

MS Mechanical Engineering

Additional Files
AlGhamaz Thesis Final Approval Form M.Alghamaz.pdf (91 kB)
Approval
Download

Share

COinS