ReMind Badge
An AI-Powered Wearable Assistant for Alzheimer's Patients
Background
Watching my grandmother struggle with Alzheimer's disease made me realise how something as simple as recognising a familiar face could become a daily challenge. I wanted to use technology to help people experiencing memory loss regain some independence and confidence in their everyday lives. This inspired me to develop ReMind Badge.
About the Project
ReMind Badge is a wearable device that uses artificial intelligence to recognise familiar faces and everyday objects, providing audio reminders to help users identify them.
Development and Engineering
I designed and built the prototype from the ground up, integrating hardware, machine learning, and product design.
-
Machine Learning: Collected approximately 200–300 training images per class and trained CNN-based recognition models using Google Colab.
-
Embedded Computing: Integrated the trained models with a Raspberry Pi and camera module, enabling the wearable to process visual inputs and generate audio reminders.
-
Software Development: Developed a Python-based dashboard to manage recognition data and support the device's functionality.
-
Product Design: Designed a custom enclosure using Autodesk Fusion 360, converted the CAD model into an STL file, and 3D-printed the casing using PLA.
-
Performance: Achieved approximately 92% recognition accuracy, with typical recognition times of 1–1.2 seconds during prototype testing.
Results and Recognition
The prototype achieved approximately 92% recognition accuracy, identifying people and objects in around 1–1.2 seconds during testing.
The project received a CREST Gold Award and an INSEF Bronze Award (2025–26).
More than anything, ReMind Badge taught me how engineering can go beyond technical innovation to address deeply personal, real-world problems.
