GaitGuard: Smart Insole for Real Time Gait Detection and Correction

Overview

Millions of people have incorrect standing postures and gait and a majority of these people are unaware they even have this issue. Many people also have undiagnosed differences in leg lengths, flat feet, pronation, supination, bow legs, and knock knees. This affects their weight distribution during standing and walking which causes permanent muscle and hip bone damage. By implementing early diagnostic tools to identify these imbalances, patients can take corrective actions to prevent chronic, long term injuries. The engineering goal of this project is to build “Gait Guard” , a low cost smart insole system that can detect abnormal postures and gaits in real time and alert users. The system uses 4 Force sensitive resistors (two on each foot) on the heel and ball of the feet to capture pressure data. An Arduino micro controller samples the analog data and transmits them via bluetooth to a custom built python application. The python application uses a Machine learning model to classify 5 different postures: Normal standing, Left leaning standing, Right leaning standing, Normal walking, Equinus gait (toe walking). On detecting an incorrect posture, the system triggers a buzzer to send an immediate alert to the user to correct the posture. Data collection included recording pressure signals every 10 ms from the sensors for 2 minutes for each of the 5 classes to train a ML model. Testing results indicate a high classification accuracy of 99%.

Competition history

• CSEF 2026 Medicine & Physiology (Junior Division) · Entry J-15-13

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Biomedical and Health Sciences projects
• CSEF Biomedical and Health Sciences projects
• Every CSEF 2026 project