Muhammad Aiman Bin Abd Khalib Universiti Teknologi Malaysia
Maintaining good posture is important for long-term health, but many people are not aware when their
posture becomes poor during daily activities. Based on our survey, 59.1% of respondents reported frequent discomfort in the neck, shoulders, and back. To address this issue, this project focuses on developing a smart posture correction device that helps users become more aware of their posture in real time, both during everyday activities and sports. The system is powered by an XIAO ESP32C3 microcontroller and uses a combination of sensors to monitor body posture. An MPU6050 IMU tracks body orientation, a flex sensor measures spine bending, and a force-sensitive resistor (FSR) detects contact pressure. The collected data is analyzed using a machine learning approach that includes pose estimation by image processing–based classification. When poor posture is detected, the system immediately alerts the user through haptic feedback using a vibrating motor. In addition, posture data is displayed on a mobile application using Blynk, allowing users to track their progress over time. Overall, this project provides a practical and sustainable solution to help people improve their posture and reduce discomfort in daily life.