ResQ

Resq

Hope in the Darkest Hour

Inspiration

Every second matters in a disaster. By observing the delays in current rescue operations, especially in under-resourced and remote regions, I wanted to create a system that saves lives by using AI and real-time geolocation to send fast alerts and connect victims with rescue teams.

What it does

ResQ is an AI-powered disaster detection and rescue system with two core parts:

1. ML Model:

• Detects floods using real-time satellite imagery (Sentinel-1 SAR & NOAA-20 VIIRS)
• Finds humans and animals in danger from drone or satellite images
• Flags high-risk zones for rescue teams

2. Web Application

• Application is built to reduce the panic among the people by letting them track rescue operations and connect local communities and first responders
• Lets users press an SOS button to send a help request
• Captures live location and checks disaster zones
• Alerts NGOs and rescue teams instantly
• Creates a public post on its posts page with location and situation to connect communities locally
• Shows nearby shelters and tracks rescue teams on a live map

How I built it

• Python and Machine Learning for disaster and victim detection.
• Streamlit to show ML results.
• React.js, TypeScript, and Tailwind CSS for the web app.
• Supabase for the backend and data.v
• Lovable.dev to host our app.
• Satellite data from Sentinel-1 and NOAA.

Challenges I ran into

• Handling and processing large satellite image files
• Improving model accuracy while keeping results fast
• Integrating multiple systems into one smooth platform

Accomplishments that I am proud of

• Achieved 92% accuracy in detecting floods and 94% accuracy for detecting wildfires
• Users can send a distress signal in just one click
• Combined AI + real-time alerts in one seamless solution
• Designed with real impact and life-saving potential

What I learned

• Working with satellite and drone data for real-world problems
• Designing a product that is user-friendly in emergencies
• Building end-to-end systems using ML + Web Technologies

What's next for ResQ

• Add detection for tsunamis and landslides
• Build a mobile app version
• Add real-time alerts on social media platforms for auto alerts
• Collaborate with NGOs and governments for real-world deployment

Built With

• ai
• deep-learning
• disaster-response
• javascript
• ml
• python
• react
• real-time-tracking
• satellite-imagery
• sdg-11
• sdg-13
• skcit-learn
• tensorflow
• vite