کنفرانس مهندسی برق ایران

صفحه اصلی / سی و چهارمین کنفرانس بین المللی مهندسی برق

A Lightweight Smart Insole with Piezoelectric and IMU Sensors: Design, Data Collection, and Machine-Learning-Based Fatigue Detection

نویسندگان :

Nima Masoudnia¹ Ali Fahim² Fariba Bahrami³ Ehsan Maani Miandoab⁴ Mohammad Khalfe Nilsaz⁵

1- دانشگاه تهران 2- دانشگاه تهران 3- دانشگاه تهران، دانشکده مهندسی برق و کامپیوتر 4- دانشگاه تهران 5- دانشگاه تهران

کلمات کلیدی :

Smart Insole،Piezoelectric Sensors،Inertial Measurement Unit (IMU)،Fatigue Detection،Machine Learning،Biomechanics،Sports Engineering،Wireless Data Transmission

چکیده :

This study presents the design, data acquisition, and machine-learning-based fatigue detection using a smart insole equipped with piezoelectric and inertial sensors. The proposed lightweight and thin insole integrates six piezoelectric sensors for plantar pressure measurement and an inertial measurement unit (IMU) for six-axis motion tracking, capturing both linear acceleration and angular velocity. Unlike conventional laboratory-based force plates, the developed system enables continuous and wireless data acquisition during natural walking and running. Real-time data transmission is achieved through an Arduino Nano 33 IoT board using Wi-Fi communication, while a custom-developed software interface allows live visualization, labeling, and manual control of recording sessions. Experimental data were collected from twelve participants under two physical conditions non-fatigued and fatigued resulting in a total of 284 labeled datasets. Multiple machine learning algorithms, including Random Forest, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Logistic Regression, and XGBoost, were trained and evaluated to classify fatigue states based on time- and frequency-domain features extracted from both sensor types. The results demonstrated that the SVM and Logistic Regression models achieved the highest accuracy of up to 100% when combining IMU and piezoelectric data. The proposed system demonstrates a robust, low-cost, and portable solution for biomechanical motion analysis and fatigue monitoring. Its combination of wearable sensor technology and machine learning provides a promising framework for real-time performance tracking and injury prevention in sports and rehabilitation applications.