Inspiration
Have you ever looked at the dashboard of your car and wondered what all the weird glowing symbols mean? Don't worry, Michael - your automobile assistant - is here to help! We are inspired by the common issue that we all face as a society - what do those symbols on the dashboard mean!?
What it does
Michael - your automobile assistant, helps the user use the application as a hub for all of their car repair needs and information. The gist of the mobile application allows the user to scan the dashboard of their car and receive feedback on what the symbols mean and how they can fix the issue without jumping through multiple hoops. The app provides the user with future actions - detailed maintenance cycles for specific tasks along with how to approach them - this helps with maintaining the vehicle.
How we built it
To build Michael - the mobile app, we used React-Native for our frontend framework alongside Python for our server, all of which is hosted on Google Cloud.
Our state-of-the-art computer-vision model is fine-tuned using Pytorch and hosted on Google Vertex AI. The model was fine-tuned with a custom dataset and is inference using API endpoints. Everything uses MongoDB Atlas for data storage and recovery.
Challenges we ran into
A few challenges we ran into: Fine-tuning the computer vision model Using MongoDB to store our vehicle data and host the server
Accomplishments that we're proud of
Building out and fine-tuning a computer-vision model(YOLOv8)
What we learned
We learned: Fine-tuning the YOLOv8 computer vision model Creating components in React-Native Using Google Cloud for the infrastructure
What's next for Michael - The Automobile Assistant
In the future, we would like to expand the app to include home maintenance - ie HVAC.


Log in or sign up for Devpost to join the conversation.