BackBuddy

BackBuddy is a compact, wearable device designed to monitor and correct standing posture.

Utilizing a XIAO ESP32S3 microcontroller and an ADXL345 accelerometer, BackBuddy collects continuous orientation data and compares it to a trained machine learning model to identify deviations from optimal posture. When poor posture is detected, the device activates a vibration motor to deliver strong and precise haptic feedback.

Additionally, Bluetooth Low Energy (BLE) facilitates efficient data communication, enabling integration with mobile devices for setup and maintenance.

My impact

  • Sole ideator and end-to-end designer, owning hardware, firmware, and system integration

  • Implemented on-device posture classification using IMU data and a trained machine-learning model

  • Designed a closed-loop feedback system combining sensing, inference, and haptic actuation

  • Integrated BLE connectivity for setup, calibration, and mobile-device interaction

  • Engineered for wearable constraints including size, power efficiency, and user comfort

Previous

Learning-Based Quadruped Locomotion & Dynamic Manipulation