Trackstar

Inspiration

More than 50% of property crimes reported to the University of California involve bicycle theft. This project aims to address this critical issue affecting UC students by providing a discrete theft detection, prevention, and tracking system. By helping students protect their bikes, we're contributing to campus safety and justice, ensuring students can focus on their education without the financial and emotional burden of bike theft.

What it does

Trackstar consists of a hardware device and mobile app:

Hardware Device: A small, discrete device that attaches to your bike and monitors for unauthorized movement
Mobile App: Receive instant alerts when unexpected movement is detected and remotely trigger an alarm to deter thieves
Real-time Tracking: When theft is detected, the device automatically sends GPS coordinates so you can track your bike's location

How It Works

Pair Your Device: Scan a QR code to pair the hardware device with your phone
Activate Watch Mode: After parking your bike, activate "Watch" mode through the app
Get Alerts: If your bike detects unexpected movement, you'll receive a push notification from the app
Track & Deter: View your bike's real-time location on an integrated map and trigger an alarm to deter thieves

How we built it

We built Trackstar as a fullstack IoT system with three main components working together:

Hardware Development: We started with an ESP32 microcontroller and integrated sensors for motion detection (MPU-6050 accelerometer) and GPS tracking (GT U7 module). The firmware uses Arduino and implements a state machine to manage Idle, Watch, and Theft Detected modes. We optimized power consumption by keeping GPS inactive until theft is detected, and implemented fast polling intervals for responsive command handling.

Backend Architecture: We built a Node.js/Express REST API with PostgreSQL to serve as the central communication hub. The backend authenticates both users (via JWT tokens) and devices (via device ID/secret pairs), stores device state and GPS coordinates, and manages the command queue that devices poll. We integrated Expo's push notification service to send real-time alerts to users when motion is detected.

Mobile App: We developed a React Native app using Expo for cross-platform compatibility. The app features a clean, animated UI with dark/light mode support, real-time status updates, and integrated Google Maps for GPS tracking. We implemented custom pulsing animations for status indicators and created a native map modal with a custom bike marker that pulses to match the theft detected state.

The entire system communicates via HTTP over WiFi, with the device polling for commands and the app using REST APIs, creating a responsive real-time monitoring experience.

Challenges we ran into

One of the biggest challenges we faced was integrating the hardware and software components by using HTTP requests to communicate between the device and our app. Some of the hardware we used had limited documentation, which required us to do lots of board level debugging. Building a cohesive testing structure on both hardware and software ends was also a necessary challenge we faced to verify the system's behavior.