Real-time evacuation management system

Akadenia is the core engineering partner for the Perimeter platform, responsible for building and maintaining the entire system. We developed the Perimeter mobile application (React Native with Mapbox), the web applications (Next.js), and the backend APIs (Node.js, PostgreSQL, AWS). The platform bridges AI-based evacuation planning with real-time collaborative situational awareness mapping. We integrated Amazon Bedrock to power generative AI features across the platform — auto-generating public-facing evacuation alerts, situation summaries, and incident reports from structured map data, dramatically reducing the time agencies spend drafting communications during active emergencies. We built a MATSim-based traffic simulation engine into the Estimator API (FastAPI/Python) for agent-based evacuation modeling — simulating vehicle flow, congestion, and clearance times on evacuation routes. We also developed an AI-powered day and night population estimation system using RasterIO and GeoTIFF geospatial data, enabling accurate population density calculations by time of day for better evacuation planning and resource allocation. Key capabilities we built include: the public-facing 'no-app' Perimeter Map (zero-friction, shareable link, no download needed), AI-powered content generation for real-time emergency communications, MATSim multi-agent traffic simulation for evacuation modeling, day/night population estimation using geospatial raster data, the full incident lifecycle from planning through execution to recovery, FEMA reimbursement automation and after-action replay, bi-directional Esri integration with KML/GeoJSON exports, and multi-agency interoperability supporting all hazards — wildfire, flood, hurricane, tornado, active assailant, hazmat, civil unrest, and planned events.