Skip to content
Back to Projects

Disaster Coordination Network — Real-Time Disaster Management System

A real-time disaster coordination platform with PostGIS-based geospatial data, DBSCAN clustering, and AI-powered vehicle recommendations using MCDM algorithms.

Python FastAPI PostgreSQL PostGIS SQLAlchemy JWT Scikit-learn Docker

Problem

During disasters, lack of coordination, resource waste, and slow response times reduce life-saving capacity. A real-time, geospatial data-based coordination system was needed.

Approach

Worked as Backend Developer in a 4-person team:

  • Geospatial Infrastructure: Location-based disaster reporting and querying with PostgreSQL + PostGIS
  • Data Integration: AFAD and Kandilli Observatory API integrations
  • AI Algorithms: Automatic task packaging with DBSCAN clustering; real-time ETA calculation with Haversine formula
  • Decision Support: AI-powered vehicle recommendation using Multi-Criteria Decision Making (MCDM)
  • Security: JWT-based authentication, Role-Based Access Control (RBAC)
  • Deployment: Container-based deployment with Docker Compose

Outcome

A scalable system equipped with AI-driven decision mechanisms enabling disaster coordinators to manage real-time location, resource, and response data.