Efficient vehicle scheduling is critical to improving customer satisfaction and optimizing resource utilization in urban environments. Inspired by the challenges of real-world routing in busy cities, we set out to create a tool that ensures fast and reliable vehicle scheduling using advanced computational algorithms.
UNDER simplifies the passenger pick up and delivery to the Vehicle Routing Problem (VRP), which is then solved using the highly optimized CP-SAT solver.
- Algorithm with very good results (more than 30% better than random allocation)
- Scalable to over 50 taxis
- For huge fleets, UNDER uses a still very good (over 20% improvement) greedy approximation algorithm
- The optimization target can be changed even during the day, to prioritise speed, energy consumption or both
We combined cutting-edge technologies and tools to bring UNDER to life:
- Python: For backend logic and CP-SAT implementation.
- OR-Tools: To handle constraint-solving and optimization.
- Socket.IO: For real-time communication between the backend and frontend.
- React: To create an interactive and responsive user interface.
- OpenRouteService: For routing accurate to the street.
- Leaflet: To visualize maps and vehicle paths seamlessly.
- Speed and Reliability: Our approach still has very good results for huge amounts of taxis.
- Scalable Design: Built a flexible system capable of handling diverse routing challenges.
- User-Friendly Interface: Developed an intuitive interface for seamless interaction and visualization.
Made with ❤️ for smarter urban navigation 🚗


