Bus Positioning System

A system that reduced average passenger wait time by 34% — built because waiting for a bus that's "5 minutes away" for 20 minutes is a special kind of frustration.

The Story

This started as a class project, but I couldn't stop thinking about it. The existing bus tracking systems were either inaccurate, confusing, or both. I wanted to build something that actually helped people make decisions — should I wait, or should I walk?

We simulated UART communication with the COTA API to get real-time data, then ran A/B tests on different UI approaches to figure out what actually reduced perceived wait time. Turns out, the way you present information matters as much as the information itself.

What I Learned

• User testing reveals things you never considered (people don't read instructions)
• A/B testing is humbling — my "obviously better" design lost
• Real-time systems require thinking about failure modes constantly
• 34% improvement sounds small until you multiply it by thousands of daily riders

Key Features

• Developed a microcontroller-based bus positioning system that reduced average passenger wait time by 34%
• Simulated UART communication between the device and COTA API to dynamically update arrival predictions
• Enhanced public safety for over 12,000 daily commuters by providing emergency-service contact through the interface
• Consulted riders and faculty to run A/B tests and refine UI usability