✅ Inspiration
The Philippines is one of the most disaster-prone countries in the world, facing frequent typhoons, earthquakes, fires, and floods. However, most people—especially those in rural or underserved communities—lack access to real-time, localized alerts or safety guidance. This inspired us to build ResQintel, an AI-powered app designed to keep Filipinos informed, prepared, and safe before, during, and after disasters.
🚀 What it does
ResQintel is a mobile application that:
- Detects fires using AI-powered image recognition.
- Tracks typhoons and displays real-time affected areas using geolocation.
- Offers earthquake preparedness guides and post-disaster checklists.
- Sends automated, localized emergency alerts and notifications.
- Provides multi-language support for inclusivity.
- Equips users with age-appropriate educational disaster modules.
🛠️ How we built it
- Frontend: Built using Flutter for seamless cross-platform mobile performance.
- Backend: Leveraged Firebase for authentication, real-time database, and cloud storage.
- AI Integration: Utilized TensorFlow and YOLOv11 for fire detection from image data.
- Data Sources: Pulled datasets from Kaggle and integrated APIs like Google Maps API, Text Recognition, and Gemini/Gemma.
- Cloud Infrastructure: Hosted and scaled using Google Cloud Platform.
🧩 Challenges we ran into
- Training the fire detection model with high accuracy using limited datasets.
- Sourcing clean, diverse, and localized disaster-related datasets relevant to the Philippines.
- Designing an intuitive UI for both tech-savvy and non-tech-savvy users.
- Ensuring reliable real-time alerting even in poor internet conditions.
🏆 Accomplishments that we're proud of
- Successfully implemented AI-powered fire detection with promising accuracy during tests.
- Created a working prototype that integrates real-time typhoon tracking and alert delivery.
- Built a multilingual interface to serve diverse Filipino communities.
- Formulated a modular education system with tailored content for different age groups.
📚 What we learned
- How to train and integrate YOLOv11 models into a mobile app using Flutter.
- The importance of user-centric design, especially in emergency apps.
- How real-time data and AI can save lives with the right integration.
- Collaboration and agile planning greatly enhanced our team productivity.
🔮 What's next for ResQintel
- Partner with local agencies (e.g., NDRRMC, LGUs) to link alerts with official sources.
- Add offline support for areas with poor connectivity.
- Expand AI capabilities to detect flooding or structural damage.
- Launch public beta testing to gather feedback and improve usability.
- Add chatbot functionality for guided disaster assistance using Gemini.
Log in or sign up for Devpost to join the conversation.