Staying Alert: Designing an AI-Driven Flood Intelligence Platform for Filipino Communities
Alerto.ai is a centralized, AI-driven nationwide flood intelligence platform designed to visualize real-time flood risks, understand contributing factors, and prepare for potential disasters. Won 2nd place out of 17 teams for the 42nd Computer Science Week Warframes Web Design Competition.
Abstract
The Philippines ranked as the most disaster-prone country in the world from 2022 to 2025 according to the 2025 World Risk Report. Every year, millions of Filipinos are displaced by typhoons and floods, yet many communities continue to struggle with limited real-time alerts. Alerto.ai addresses this gap as a centralized, AI-driven intelligence platform powered by Project NOAH. The prototype consists of a nationwide 3D flood hazard map, an AI-assisted chatbot, and text alerts. It allows residents to visualize flood risks, query safety instructions, locate nearby resources, and receive synchronized announcements coordinated directly with local government units.
Problem Statement & Background
The Challenge
How might we design a web app prototype that uses AI/ML to help communities visualize flood risks, understand contributing factors, and prepare for potential disasters?
The Philippines holds a title that no nation wants: from 2022 to 2025, it consistently ranked as the most disaster-prone country in the world according to the 2025 World Risk Report. Every year, millions of Filipinos are displaced by typhoons and floods, yet many communities continue to struggle with a critical lack of real-time alerts and very slow disaster responses.
While billions of pesos are allocated to flood control projects, a severe lack of transparency and implementation has left these communities even more vulnerable. Filipinos deserve more than uncertainty.
Alerto.ai was built as a 3-in-1 compiled solution for information dissemination, integrating a nationwide 3D map, an AI-assisted chatbot, and text alerts to shift communities from reactive disaster response to proactive flood preparedness.
Design Strategy
Nationwide 3D Interactive Map: Visualizes streets prone to floods and precise water depths based on the shade of blue (the darker the blue, the higher its severity is). A central information box details the nearest flooded area and contributing factors (e.g. heavy rainfall, nearby body of water). Filter buttons allow users to toggle overlays for nearest evacuation centers, nearby supply stores, and emergency contacts.
Alerto Chatbot: A lightweight chatbot interface designed for users with low-bandwidth connections who cannot easily load heavy map visuals. Users can query disaster preparedness tips, evacuation centers, nearby stores, flood-prone areas, and go-bag essentials.
Free Text Alerts: Free SMS subscription where users input their phone number. A text message is sent to confirm their status, and they can converse with the Alerto chatbot via SMS as if they were online. The AI processes incoming locations to identify users in high-risk zones and prioritizes automated text alerts to them first.
Typography & Colors
Methodology
- Designed a responsive desktop dashboard offering a nationwide 3D map, nearby supply search, and local hotline directory.
- Created a chatbot feature to understand conversational queries and provide relevant information based on the user's location and the current situation.
- Incorporated predictive modeling telemetry from Project NOAH data to predict real-time water pooling and road passability.
Results & Discussion
Alerto.ai successfully maps flood severity and coordinates emergency resources using predictive analytics, natural language processing, and location-aware geofencing.
| Features | Key Component | Solution |
|---|---|---|
| Predictive Map Mapping | Predictive Modeling | Analyzes historical rainfall and terrain to predict exactly how water will pool in specific streets before it happens. |
| Spatial Navigation | Object Detection & Routing | Processes 3D spatial data to identify passable vs. impassable roads, ensuring the suggested path is safe to travel. |
| Barangay geofencing | NLP & Geofencing | Scans user coordinates to prioritize Barangay-level emergency contacts first before escalating to LGU channels. |
| AI Web Chatbot | Natural Language Processing | Extracts user intent from queries like 'What is the flood status?' and delivers instant verified safety data. |
| Text Alert Prioritization | NLP & Priority Queueing | Processes location texts, maps users in high-risk zones, and alerts them first to ensure timely notifications. |
The system's predictive routing and geofencing ensure that both local authorities and citizens are synchronized, reducing dispatch delay during emergency scenarios.
With a 3-in-1 multi-channel framework, Alerto.ai ensures information is accessible to users with high-speed internet, mobile web data, or basic SMS network coverage.
Conclusion & Next Steps
Alerto.ai represents a vital shift from reactive disaster responses to proactive, data-driven approaches.
By applying AI to flood predictions and local government coordination, the platform ensures that every Filipino can receive real-time alerts and actionable safety information regardless of internet connectivity.
Alerto.ai sets a benchmark on how the Philippines can build resilience, using information to mitigate damage, innovation to expand capabilities, and early warnings to save lives.
Skills & Frameworks
Explore Other Case Studies
Streamlining the Workflow: Evaluation of InShot and FastSnap as a Prototype in Optimizing Video Editing Experience of College Students
FastSnap is a mobile video editor prototype made using Figma. It is designed to offer an intuitive user experience for users who edit videos on the go.
Establishing a Reliable Marketplace Platform for Cosplay Products in the Philippine Community
CosMarket is a mobile e-commerce application prototype designed to revolutionize the cosplay community by offering a seamless, secure platform for buying, selling, and renting cosplay outfits and accessories.
Design and Viability of Payonnect as a Sustainable Smart Umbrella Sharing Kiosk Network
Payonnect is an eco-friendly mobile application and kiosk sharing prototype that allows commuters to borrow and return umbrellas through QR codes, providing immediate shelter during unexpected weather.