Dynamic Route Planner

Abstract

This paper presents a web-based dynamic route planning application that integrates real-time traffic and weather data into path computation. The system is implemented with a modern web stack: React and TypeScript on the front end (via Vite and Tailwind CSS for rapid UI development) and Node.js with Express.js and MongoDB on the back end. It computes single-destination routes using classic shortest path algorithms (Dijkstra’s and A* search) and solves multi-stop (travelling salesman) routing using the Held–Karp dynamic programming algorithm. Real-time traffic information is fetched via the Google Maps Platform, and current weather conditions are obtained from the Open Weather API. These data sources are used to dynamically weight road segments, so that routes avoid congested or hazardous areas. In testing scenarios, the planner yields significantly faster or safer routes than static baselines (e.g. reducing travel time by leveraging up-to-date traffic flows). Complexity analysis confirms that the Held–Karp solver is optimal but exponential in the number of stops (O(n2 2n)), so the system scales to moderate-sized tours. Overall, this work demonstrates how integrating live data with established graph algorithms can improve route guidance in complex urban environments.