Visual Twin: Enhanced Railway Asset Mangment

This repository hosts the Visual Twin Open Data project proposal, a visual data-driven railway monitoring platform. It integrates a Streamlit app, interactive visualisation, object detection, and chatbot interaction for enhanced railway asset condition monitoring and supporting predictive maintenance. Play with the app directly - Railway Visual Twin v2.0, or follow the instructions below to DIY it for more fun.

Features

• Visual Railway Monitoring: Interactive visualization of UK railway usage and routes.
• Chatbot Interaction: Integrated chatbot for inquiries.
• Object Detection Module: Detects specific conditions in railway assets.

Installation

1. Clone this repository:

   git clone https://github.com/yjq349825834/VisualTwin.git
   cd VisualTwin
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Run the Streamlit app:

   streamlit run ./main_app.py
   

Suggested Sources of Open Data

• Open Sensor Data for Rail 2023
• The Office of Rail and Road Data Portal
• Network Rail Data Feeds

Project Structure

VisualTwin/
│
├── main_app.py                    # The main application file
├── module_chatbot.py              # The chatbot module
├── Dockerfile                     # Dockerfile for containerizing the app
├── requirements.txt               # Python dependencies for the app
├── README.md                      # Project description and instructions
├── data/                          # Folder for datasets or statics
│   ├── route_info_vibrations.csv  # Route info between London Kings Cross and Cambridge for demo
│   └── videos.mp4                 # Demo videos
│   └── images.png                 # Demo sensor data
    └── data.ods                   # Example dataset from ORR (Train Station Usuage)
├── notebooks/                     # Jupyter notebooks for instructions or experiments
│   └── object_detection.ipynb     # Instruction notebook for object detection module