ResistX: AI-Powered Decentralized Resilience Platform
During major natural disasters or climate hazards, centralized communication infrastructures and internet connectivity are often the first things to fail. This leaves rescue teams in the dark and trapped individuals disconnected from immediate help.
To solve this critical vulnerability, I built ResistX during Hackhazards '26—an offline-first, decentralized disaster response platform designed to operate completely off-grid.
The Core Technical Architecture
ResistX is built on a robust, highly responsive tech stack optimized for extreme low-bandwidth and offline environments:
- Frontend Dashboard: Built using React.js and Tailwind CSS for a highly tactical, responsive dark-themed command interface.
- Backend & Processing APIs: Powered by Python and FastAPI to manage localized data streaming.
- Offline Vision-Language Models (VLM): Simulates processing drone and satellite feeds directly on edge computing nodes (like Jetson Nano or Raspberry Pi) to identify structural debris, flooding levels, and stranded people without internet dependencies.
- Decentralized Communications: Aggregates off-grid emergency SOS data from the ground using localized Bluetooth, Wi-Fi, and LoRa mesh networks.
How It Works
- Data Acquisition: The system intercepts visual feeds locally. An offline edge computer evaluates the images to construct a dynamic, localized "Hazard Heatmap".
- Mesh Aggregation: Stranded community members broadcast lightweight distress signals from their mobile devices over local mesh nodes.
- Tactical Action: Emergency management units view a unified dashboard showing exactly where trapped clusters are located and immediately map out optimal, offline evacuation routing.
By moving computation to the edge and infrastructure to local mesh nodes, ResistX ensures that saving lives doesn't depend on an internet connection.
Top comments (0)