AI Gesture Recognition Device

Skills Used: Python, C++, Swift, TensorFlow, Soldering, Fusion 360, CAD

I developed a wearable, gesture-controlled reminder device designed to help children and adults with Autism and other special needs manage their daily tasks and routines—fostering greater independence and reducing harmful repetitive behaviors. The system integrated a Raspberry Pi Zero 2 W, a PAJ7620U2 gesture sensor, and audio playback components, all enclosed in a custom 3D-printed case built for portability and durability.

To prioritize accessibility, the device was fully controlled through gesture recognition ( as seen on the right ) , eliminating the need for screens or buttons. It paired seamlessly with WaveLink, a companion iOS and macOS app I built in Swift, which enabled caregivers and teachers to record personalized audio tasks and wirelessly send them to the device via Bluetooth or Wi-Fi hotspot.

On the embedded side, I developed the core logic in Python and C++ to handle gesture recognition, task scheduling, and real-time synchronization. The project emphasized affordability, usability, and privacy, and incorporated user feedback from human subject testing to improve its functionality and reliability.

PROJECT LINKS:

- Senior Capstone Presentation

- Published Thesis

CAD Model
App Screenshots
App Screenshots
LinkedIn Email Instagram GitHub