ResQLink AI – Smart Emergency Response Network

Bridging Witnesses, Drivers, and Hospitals for Real-Time, Life-Saving Actions

Inspiration

Every second matters in a life-threatening accident. We realized:

50% of trauma-related deaths occur within the golden hour.
40% of witnesses hesitate or fail to report accidents effectively.
30% of ER resources are misallocated due to unstructured pre-arrival data.

ResQLink AI empowers witnesses, first responders, and hospitals with AI-powered tools for real-time, structured, and intelligent emergency response.

What It Does

ResQLink AI is an end-to-end, AI-powered accident response enhancement system that:

Allows witnesses to report accidents via voice, text, or image using a SOS-enabled interface.
Uses NLP to extract injury-related keywords from descriptions.
Applies Computer Vision to detect injury severity and affected body parts from images.
Integrates Google Gemini AI to:
- Suggest nearest hospitals
- Recommend real-time first aid
Notifies hospitals in advance with AI-generated triage to prepare appropriate care.
Driver Assignment System:
- Assigns the nearest available driver using a min-heap queue.
- Displays shortest route to the accident and to hospitals using Dijkstra's Algorithm.
- Sends first-aid guidelines and injury insights to the assigned ambulance driver.

System Architecture

Base Architecture Diagram

System Architecture

Description

The diagram above illustrates the end-to-end architecture of the ResQLink AI system:

SOS Alert Trigger: A user initiates an alert via voice, text, or image.
Timestamp and Location are captured immediately and routed to the backend.
Image Processing: If an image is submitted, it is passed through a CNN-based Computer Vision model to assess the severity of the accident.
Text/Voice Processing: Textual or voice input is analyzed using NLP models to predict the potential body parts injured.
Backend API: All inputs are processed by a backend system built using Express/Node.js.
Data Handling: All relevant data is stored and managed in MongoDB.
Google Gemini AI Integration:
- Suggests the nearest hospital based on injury severity and geolocation.
- Provides first-aid recommendations tailored to the reported injuries and severity.

Key AI Features

Capability	Technology Used
Injury NLP Extraction	Custom NLP Pipelines
Image-based Diagnosis	Computer Vision (CNN)
Hospital/First Aid	Google Gemini AI
Route Optimization	Dijkstra’s Algorithm
Driver Assignment	Min-Heap Priority Queue

Tech Stack

Component	Tools & Technologies
Frontend	React.js, Tailwind CSS
Backend	Flask (AI), Node.js + Express.js (API)
AI & ML	NLP, Computer Vision, Google Gemini AI
Infrastructure	Docker, GitHub, Cloud Deployment

How We Built It

Developed frontend using React and Tailwind CSS for a clean UX/UI.
Built modular backend using Flask (Python) and Node.js/Express.
Trained and deployed NLP and Vision AI models.
Integrated Google Gemini AI for hospital and first-aid recommendations.
Dockerized microservices for scalable cloud deployment.
Designed ambulance driver interface using Dijkstra’s algorithm for route optimization.

Challenges Faced

Processing inconsistent or incomplete witness reports.
Training accurate image classifiers for injury severity detection.
Reducing latency in AI processing and Google Gemini integration.
Handling high traffic and ensuring real-time scalability.
Lack of time to design a fully interactive UI.

Accomplishments

Created a full-stack AI-powered emergency response system.
Integrated NLP, Computer Vision, and Google Gemini AI.
Implemented AI-based hospital alerting and triage.
Used Dijkstra’s Algorithm and min-heap queue for driver routing.
Built scalable, cloud-deployable microservices using Docker.

What We Learned

The importance of structured reporting in emergency response.
The role of AI in transforming hospital coordination and triage.
The potential of NLP and Computer Vision in real-world accident detection.
Scalability challenges when handling real-time emergency data.
How Google Gemini AI can enhance emergency recommendations.

What’s Next for ResQLink AI

Add live video analysis using extended Computer Vision models.
Enable multi-language support for NLP pipelines.
Integrate IoT data from vehicles and wearables for automated accident detection.
Use blockchain to secure and track emergency incident records.
Implement predictive AI models to identify accident-prone areas.