Smart Transit Hub: Real-Time Commuter Forecasting

Smart Transit Hub is a one-day hackathon prototype that helps Filipino commuters avoid blind waiting at public transport terminals.

The live demo focuses on PITX Gate 3 during rush hour. The system reads a terminal camera feed, counts visible commuters, estimates the current waiting time, and recommends the next best action for commuters and terminal staff.

Problem

Public transport congestion makes opportunity inequality worse. A student can miss class, a worker can lose paid time, and a family can lose hours simply because they had no clear view of how long a terminal queue would take.

Solution

Smart Transit Hub gives commuters and terminal teams three simple answers:

1. How many people are visibly waiting?
2. How long might the wait be?
3. What should people do next?

The prototype supports SDGs 4, 8, 10, and 11 by improving access to education, protecting work hours, reducing mobility inequality, and helping cities manage transport spaces better.

Demo Features

• PITX commuter page with live video stream
• Anonymous person counting from the video feed
• Explainable wait-time estimate
• Route or terminal action recommendation
• Operations dashboard for terminal staff and LGU-style decision makers
• Fallback demo mode when the sample video or YOLO model is not yet available

Tech Stack

• Frontend: HTML, JavaScript, Tailwind CSS CDN
• Backend: Python Flask
• Video processing: OpenCV
• Person detection: Ultralytics YOLOv11 Nano when yolo11n.pt is available
• Forecasting: lightweight explainable queue model

Folder Structure

smart_transit_hub/
├── app.py
├── forecasting_ai.py
├── people_counter.py
├── route_ai.py
├── requirements.txt
├── yolo11n.pt                 # optional, place here
├── data/
│   └── pitx_rush_hour.mp4      # optional, place here
├── templates/
│   ├── base.html
│   ├── index.html
│   ├── commuter.html
│   └── lgu_dashboard.html
└── static/
    ├── css/style.css
    └── js/
        ├── home.js
        ├── commuter.js
        └── dashboard.js

How to Run

python -m venv .venv
.venv\Scripts\activate     # Windows
pip install -r requirements.txt
python app.py

Open:

• http://127.0.0.1:5001/
• http://127.0.0.1:5001/commuter
• http://127.0.0.1:5001/dashboard

Important Demo Note

If data/pitx_rush_hour.mp4 or yolo11n.pt is missing, the app still runs using a visual fallback feed. This prevents the pitch from failing during setup.

Privacy Promise

The prototype only uses anonymous counts. It does not save videos, faces, names, or personal identities.