English
Français
Divia Transport Planner
The Divia Transport Planner project is built on a simple and efficient architecture, designed to leverage open data from the Divia public transport network in Dijon. The application queries public APIs to retrieve timetables and routes, processes them with Python and Pandas, and provides optimized journey suggestions. The interface offering an interactive and user-friendly experience. This clear separation between data collection, business logic, and presentation ensures maintainability and scalability.
Architecture
- Python: main language for data collection and processing.
- NetworkX: library for graph modeling and analysis, used to optimize routes.
- Gradio: interactive web interface for quickly testing and deploying models and features.
- Hugging Face: integration of pre-trained models to enrich the application with advanced AI capabilities.
- GTFS: standard format for public transport data, used to import and process timetables and routes.
- FastAPI: high-performance framework for building RESTful APIs, enabling integration with external services and scalable deployment.
Divia Transport Planner is a Python application designed to help users plan their journeys on Dijon's public transport network. It simplifies the search for timetables and routes by directly leveraging Divia open data. The application also integrates a chatbot capable of answering frequently asked questions from users, making the experience more interactive and accessible. The goal is to facilitate urban mobility and improve user support through a clear and intelligent interface.
Many users struggle to quickly find the best routes or timetables for public transport. Existing tools are sometimes complex or not well adapted to everyday needs. The challenge was to design a simple and intelligent solution capable of guiding users to the most relevant journeys.
The project includes:
- Simplified access to Divia open data.
- Route modeling and optimization using graphs and the NetworkX library.
- An interactive interface with an integrated chatbot.
- A scalable solution adaptable to user needs.
Key Features
- Use of Divia open data (GTFS)
- Route optimization with graphs and NetworkX
- Interactive interface with Gradio or Streamlit
- Chatbot integration for frequently asked questions
- Use of FastAPI to expose RESTful APIs
- Extensible with Hugging Face to enhance AI