Enhancing Sitting Posture Detection and Alert System with Multi-Sensor Integration

Poor sitting posture among office workers is a common issue that contributes to musculoskeletal disorders, including low back pain. This paper presents an enhanced sitting posture detection and alert system based on multi-sensor integration. The system incorporates flex sensor for the thoracic spine and MPU6050 sensor for the lumbar spine within a wearable vest to precisely monitor spinal angles. A vibrating motor provides real-time haptic feedback upon detecting poor posture, while data processing is managed by an ESP32 microcontroller, with results displayed on the Blynk application. Performance evaluation, conducted through expert consultation and data collection from ten participants, demonstrated that the multi-sensor system achieved 100% accuracy in detecting good posture, eliminating misclassification errors. In contrast, the single-sensor system attained <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{8 0 \%}$</tex> accuracy, misclassifying <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{2 0 \%}$</tex> of poor postures. Notably, the MPU6050 sensor significantly enhanced detection reliability, particularly in identifying poor posture cases. These findings indicate that the proposed multi-sensor approach improves detection accuracy by 20 % over single-sensor systems, offering a promising solution for enhancing posture awareness and reducing posture-related health risks in workplace environments.