RG.

The Inaugural 2025 Durham College Hackathon

🏆 My team placed 5/12 at Durham College's inaugural 24-hour hackathon (sponsored by DRT) with a React Native app called Driver Dashboard.

TL;DR:

In just 24 hours, we developed Driver Dashboard, a cross-platform mobile app that improves communication between bus drivers and riders. It provides real-time bus updates for riders and enables drivers to report delays or emergencies. Built with React Native, Python, JavaScript, and machine learning, and networked using Ngrok, the project earned us a top 5 finish and was an incredible learning experience in rapid development, applied AI, and teamwork.

Introduction

In early 2025, Durham College hosted its inaugural 24-hour hackathon — a high-intensity sprint of innovation and problem-solving. Students across programs came together to ideate, build, and deliver working tech solutions within a single day. It was more than just a coding challenge; it tested creativity, communication, and resilience. I was part of a three-member team of developers and designers. Our mission: build a functional app that could address real challenges faced by the Durham Regional Transit (DRT) system.

Conceptualizing Driver Dashboard

Public transit is often disrupted by unforeseen events, but communication with passengers is rarely timely. To tackle this, we created Driver Dashboard, a mobile app with two distinct user roles:

Drivers

Drivers could report:

  • Traffic delays
  • Road closures
  • Emergencies (e.g., mechanical failures, safety issues)
    — all directly from their interface in the app.

Riders

Riders could:

  • View live updates on their bus
  • Track expected arrival times
  • Visualize their bus on a live map within a certain radius

This two-way communication system aimed to boost rider confidence and help drivers proactively manage disruptions.

Technical Architecture and Tools

Choosing the right stack was essential given our time limit. Here’s what we used:

React Native

  • Enabled fast development of a cross-platform mobile interface
  • Supported both Android and iOS deployment from a single codebase

Python

  • Handled backend logic and API data processing
  • Managed server-side operations, data flow, and integration with ML models
  • Processed data from the Durham Regional Transit API

Machine Learning

  • Used a basic ML model to analyze real-time and historical data
  • Predicted arrival delays and enhanced ETA accuracy for users

Ngrok

  • Allowed us to expose our local backend to the internet securely
  • Enabled real-time API interaction between the frontend and backend

We also used Git for version control and online collaboration tools to stay organized and on schedule.

Development Process

The hackathon began with a sprint planning session where we identified goals, defined the MVP, and assigned tasks based on each team member’s strengths.

  • I focused on building the rider-facing UI in React Native, ensuring that data could be consumed and visualized in real-time.
  • I also contributed to the driver input form and integrated it with backend APIs.
  • In parallel, our backend team built Python APIs that processed reports from drivers and served them to the frontend in real time.
  • We implemented a simple machine learning model to detect and predict delay patterns using time/location data.
  • Finally, we used Ngrok to solve networking issues quickly and allow real-world testing without complex infrastructure.

Despite the short timeline, we built a fully functioning prototype — and even had time for a round of testing.

Challenges and Solutions

  • Time Constraints: We had only 24 hours. By defining a clear MVP and sticking to core features, we avoided scope creep.
  • Data Accuracy: Real-time bus data is noisy and inconsistent. Our ML model helped smooth out these inconsistencies and make better predictions.
  • Integration: Connecting the frontend and backend in a live environment was tricky. Ngrok provided a seamless tunnel for rapid testing and integration.
  • Team Coordination: Coordinating across three people with limited time meant constant communication. We used frequent check-ins to stay aligned.

Key Contributions

  • Built the rider-side interface in React Native, including real-time update displays
  • Helped implement the driver input form for reporting issues
  • Developed backend scripts in Python for data ingestion and processing
  • Integrated Ngrok for live server tunneling and testing
  • Co-developed and tuned a basic ML model to support ETA predictions
  • Participated in live testing and debugging across both platforms

Lessons Learned

This hackathon pushed me out of my comfort zone and taught me a lot about:

  • Rapid prototyping: Simplicity and clarity are key when time is limited.
  • Real-world teamwork: Communication and clear ownership make all the difference.
  • ML in production: Even lightweight models can add meaningful features.
  • Technical stack choices: The right tools accelerate development and reduce friction.
  • Resilience: With focus and teamwork, even ambitious ideas can become reality in 24 hours.

Final Thoughts

Finishing in the top 5 was a proud moment — but what mattered most was the experience: learning new tech, building under pressure, and solving real-world problems with a great team. I'm excited to carry these lessons into future hackathons and software projects.