Smart Train Route Finder

India runs 13,000+ trains daily, carrying 25M+ passengers, yet booking is tough. Train Route Finder uses AI to find routes with seats, predict delays, and process voice commands.

Comment

Inspiration

My daily struggle to book train tickets in India drove the creation of Smart Train Route Finder. I often faced the frustration of never securing a confirmed seat, ending up in overcrowded general compartments where discomfort and chaos were the norm. This experience pushed me to find a solution that would make train travel more predictable and comfortable.

What it does

Smart Train Route Finder is an AI-powered system that helps passengers discover alternative routes with available seats. It leverages voice commands to capture travel requests, extracts relevant details through natural language processing, and predicts train delays—offering a streamlined, hassle-free travel planning experience.

How We Built It

Frontend: Developed using HTML, CSS, and JavaScript.
Backend: Built with Flask (Python) to handle server-side logic.
Web Scraping: Employed Selenium to scrape train schedules and seat availability.
Data Processing: Used Python to process and manage scraped data.
Version Control: Leveraged Git to track changes and manage feature additions.
AI Integration: Utilized Azure Speech-to-Text for converting voice commands and Azure NLP for extracting travel details.
GitHub Copilot: Accelerated coding, debugging, and testing with GitHub Copilot, which provided instructions and walkthroughs for faster development.
Deployment: Hosted on Google Cloud Platform (GCP) with Vertex AI to provide delay predictions.

Challenges we ran into

Optimal Route Identification: Analyzing multiple train combinations and ensuring realistic connection times proved challenging. Delay Prediction: Processing historical train data to predict delays accurately was a significant technical hurdle. Error Handling: Ensuring the system gracefully managed instances when no viable routes were found required careful debugging. Optimizing Route Speed: Predicting the best intermediate station between two given stations was complex. Rather than using distance as a parameter—which would have increased workload and waiting time—we first used a brute force approach to identify all promising intermediate stations, then trained a model on that data. This smart two-step approach balanced efficiency with accuracy.

Accomplishments that we're proud of

Successfully integrated multiple Azure AI services to deliver an end-to-end solution. Created a system that transforms the stressful experience of booking train tickets into a reliable, user-friendly process. Demonstrated real-world impact by addressing a problem faced by millions of daily train travelers in India. Empowered users to make wise travel decisions by providing alternative routes with available seats and delay predictions, ensuring optimal utilization of available capacity.

What we learned

The power of leveraging cloud-based AI services to solve real-life problems. How to convert voice inputs into actionable data using advanced NLP techniques. The importance of optimizing backend processes for handling complex, dynamic data. Best practices for deploying scalable AI solutions on cloud platforms.

What's next for Smart Train Route Finder

Future plans include expanding the system to cover more routes, refining delay predictions with additional data sources, developing a mobile application for on-the-go access, and continuously enhancing user customization options. The goal is to make train travel in India even more efficient and stress-free for everyone.

Built With

Share this project:

Updates