RoadRescue: A Real-Time AI-Powered Web Platform for Roadside Emergency Assistance and Service Provider Management

Authors:

Ayaan Khan

Department of Computer Engineering Vivekanand Education Society’s Polytechnic Mumbai, Maharashtra, India co2023.ayaan.khan@ves.ac.in

Jiya Sachdev

Department of Computer Engineering Vivekanand Education Society’s Polytechnic Mumbai, Maharashtra, India co2023.jiya.sachdev@ves.ac.in

Samad Khan

Department of Computer Engineering Vivekanand Education Society’s Polytechnic Mumbai, Maharashtra, India co2023.samad.khan@ves.ac.in

Pratibha Pednekar

Department of Computer Engineering Vivekanand Education Society’s Polytechnic Mumbai, Maharashtra, India pratibha.pednekar@ves.ac.in

Abstract—Roadside vehicle emergencies such as battery fail- ure, flat tires, fuel exhaustion, lockouts, and minor accidents pose significant safety and mobility challenges, particularly in densely populated urban areas. This paper presents RoadRescue, a real- time, AI-powered single-page web application that facil- itates rapid connection between stranded drivers and nearby service providers. The platform integrates a conversational AI chatbot powered by Google Generative AI (Gemini), real-time interactive mapping using Leaflet, and secure user authentication via Firebase. Experimental evaluation of the functional prototype demonstrates effective real-time coordination, responsive cross- device usability, and reliable AI interaction, highlighting the potential of intelligent web-based systems for modern roadside assistance.

Index Terms—Roadside Assistance, Web Application, Gener- ative AI, Real-Time Geolocation, Leaflet Mapping, Emergency Response, Single-Page Application, Firebase Authentication