Inspiration

Cars are a cornerstone of the American dream, symbolizing freedom. Yet, when a dashboard warning light flickers on, it often brings confusion and anxiety. Imagine a different scenario: instead of feeling overwhelmed by cryptic lights or uncertain about your mechanic’s recommendations, you’re in control. With complete confidence, you understand exactly what your car needs—empowering you to make informed decisions and keep your vehicle running smoothly.

What it does

With just a snap, our app identifies car lights, provides detailed descriptions, and estimates replacement costs—giving you the tools to make informed decisions and avoid unnecessary expenses. Whether you’re a car enthusiast or just want to avoid being overcharged, CarBuddy simplifies the process of understanding and servicing your car. It's not just an app—it’s your trusted companion on the road to smarter car care.

How we built it

At its core, a TensorFlow Lite model powers the backend, enabling efficient and accurate image recognition of car lights. The intuitive front end was developed using Flutter. To handle file storage and retrieval, Pinata was integrated, ensuring secure and reliable management of image data.

Challenges we ran into

The biggest challenge was getting the ML Model to pair with Pinata. There were countless tutorials on how to set up ML models with Firebase, but I want to make use of Pinata for faster file storage and retrival.

Accomplishments that we're proud of

I am extremely proud of implementing my first-ever ML model in an application. I have always wanted to experiment with AI/ML and now finally got a shot at using an image-classifying ML model that can help me identify lights and parts.

What we learned

I learned the importance of time management in implementing an app of this caliber in a short amount of time.

What's next for CarBuddy

I want to implement custom schedules so that you can track the history of your vehicle's service and know exactly what you need next for the smooth operation of your vehicle.

Built With

Share this project:

Updates