UtkarshRake: Rake Formation Optimiser

UtkarshRake—where 'Utkarsh' means progress—is an AI-powered tool that optimizes rake formation to enhance rail logistics, boosting efficiency, cutting emissions, and reducing waste.

Comment

Inspiration

The idea for UtkarshRake emerged during the Smart India Hackathon, where I encountered the challenge of optimizing rake formation in railway logistics. Inspired by the potential to improve economic efficiency and reduce environmental impact, I set out to build a solution using machine learning and operations research.

What It Does

UtkarshRake is an AI-powered decision support system that automates rake formation by analyzing demand orders, wagon availability, and yard constraints. It intelligently matches wagons to cargo needs, optimizing logistics while reducing CO₂ emissions and minimizing waste.

How We Built It

  • Developed a frontend using Streamlit with a polished UI and sidebar navigation (Home, Data Upload, View Data, Optimization, Analytics)
  • Implemented optimization logic using PuLP for constraint-based planning
  • Integrated a machine learning model to predict demand categories
  • Added real-time analytics to monitor wagon utilization and order distribution
  • Designed modular components for scalability and future ERP integration

Challenges We Faced

  • Handling inconsistent CSV formats and missing data entries
  • Designing a flexible optimization model adaptable to varying demand and wagon types
  • Ensuring performance and responsiveness across large datasets

Accomplishments We're Proud Of

  • Built a working prototype that demonstrates intelligent rake formation
  • Successfully combined ML-based demand prediction with optimization planning
  • Created a user-friendly, navigable interface for logistics personnel
  • Visualized key metrics to support data-driven decisions in freight operations

What We Learned

  • Gained hands-on experience with operations research and constraint programming
  • Learned to integrate ML and optimization for real-world decision-making
  • Understood the complexities of logistics planning and the value of automation

What's Next for UtkarshRake

  • Add visualization tools for rake layout and performance metrics
  • Build a robust backend for real-time data ingestion and ERP integration
  • Expand the solution to other industries where logistics optimization is critical
  • Explore predictive maintenance and route optimization using advanced ML models

Built With

  • or-tools
  • pulp
  • python
  • streamlit
  • tensorflow
Share this project:

Updates