Visture

The Project

Postural dysfunction refers to critical body posture that develops from prolonged poor sitting habits and constant pressure on the body. 22,6 million people in Indonesia suffer from postural dysfunction in 2023. Postural dysfunction not only leads to long-term physical impairment, but also negatively affects mental well-being and productivity. This also reflects rising demands for postural dysfunction major treatments: physiotherapy and back braces. However, physiotherapy is expensive and is costly to Indonesian standards, while back braces causes skin irritation and only covers half-body. In school, I’ve seen classmates and even teachers suffer postural dysfunction due to poor ergonomic habits from prolonged sitting often without realizing it. Visture is the first real-time ergonomic intelligence platform that uses both computer vision (AI) and gyroscopic sensors to accurately detect and correct postural dysfunction. Visture supports long-term physical health through continuous subtle vibration reminders that gradually build consistent posture muscle memory with several key innovations: 1. Hybrid Posture Detection System: Integrates computer vision (AI) through a smartphone camera with wearable gyroscopic sensors to provide accurate full-body posture monitoring. 2. Real-Time Muscle Memory Development: Subtle vibration feedback from wearable sensors and smartphone alerts encourages immediate posture correction. 3. Progress Tracking, Detailed Analysis: Provides visual performance insights and posture improvement metrics through an intuitive and engaging mobile interface. 4. Low-profile Design: Lightweight wearable sensors & phonesized AI model allow continuous full-body correction in everyday environments (e.g schools, offices) without complex setup. ----- Hardware Specifications: - MPU6050: A 6-axis inertial measurement unit (IMU) containing a gyroscope and accelerometer, used to detect body orientation, tilt, and movement for lower-body posture monitoring. - ESP32: Microcontroller used to process posture data collected from sensors and transmit the data to the mobile application via Bluetooth. The microcontroller provides sufficient processing capability, integrated wireless connectivity, and low power consumption. - GOOJODOQ 4000mAh: Provides portable electrical power for the wearable hardware components. Combining three ESP32 modules alongside a phonesized posture detection model inside the mobile app is difficult. I tried using many frameworks and they failed, but I switched to Flutter and I was able to integrate everything inside. Designing the gyrosensors also need to be as lightweight and compact for better comfort aswell.

View this project on YouTube