How rajatrawal-06/floodapp works
FLUX combines a mobile disaster-response dashboard with a FastAPI backend for satellite flood detection, geospatial hazard analysis, evacuation planning, reports, alerts, and conversational assistance. The main workflow is a user-initiated detection or location query, followed by flood inference and physical-context enrichment; hazards can then inform triage and evacuation routes shown to responders. The README documents the broader capability flow, while sampled excerpts confirm selected routing, triage, inference, and API behavior; unsampled subsystem wiring is represented only where documented.
Main entry points: App.tsx Mobile app · endpoints.py HTTP API · evacuation_engine.py Evacuation planner · flood_agent.py Flood agent
Components 26
Mobile dashboard
- Mobile app
dell-flood-app/App.tsx - Map view
dell-flood-app/src/components/MapView.tsx - Crowd reports
dell-flood-app/src/components/ReportForm.tsx - Triage panel
dell-flood-app/src/components/TriagePanel.tsx - Chat panel
dell-flood-app/src/components/ChatbotPanel.tsx - API client
dell-flood-app/src/services/api.ts
Backend service
- HTTP API
dell-flood-backend/app/api/endpoints.py
Hazard analysis
- Flood inference
dell-flood-backend/app/ml/inference.py - Mask postprocessing
dell-flood-backend/app/ml/postprocess.py - MirEye physical data
dell-flood-backend/app/flux/api/client.py - TWI and susceptibility
dell-flood-backend/app/flux/pipeline/physical_data_hazard_engine.py - Predictive hazards
dell-flood-backend/app/flux/pipeline/flood_prediction_pipeline.py - Population clusters
dell-flood-backend/app/flux/pipeline/cluster.py
Response planning
- Priority and allocation
dell-flood-backend/app/flux/pipeline/triage_engine.py - Evacuation planner
dell-flood-backend/app/flux/routing/evacuation_engine.py - Elevation-aware routes
dell-flood-backend/app/flux/routing/graph_router.py - Shelter discovery
dell-flood-backend/app/flux/shelters/finder.py - Tactical maps
dell-flood-backend/app/flux/visualization/map_plotter.py
AI assistance
- Flood agent
dell-flood-backend/app/agent/flood_agent.py - MirEye chatbot
dell-flood-backend/app/agent/mireye_chatbot.py - Agent tools
dell-flood-backend/app/agent/tools.py - RAG report generator
dell-flood-backend/app/RAG/generator.py - Report knowledge store
dell-flood-backend/app/RAG/vector_store.py
Other components
- Responder or resident
- MirEye API
- LLM provider
Connections 25
- Responder or resident → Mobile app: uses
- Mobile app → API client: requests
- API client → HTTP API: calls
- Mobile app → Map view: displays
- Mobile app → Crowd reports: presents
- Mobile app → Triage panel: presents
- Mobile app → Chat panel: presents
- HTTP API → Flood inference: runs inference
- Flood inference → Mask postprocessing: produces mask
- TWI and susceptibility → MirEye physical data: enriches context
- MirEye physical data → MirEye API: fetches data
- Predictive hazards → TWI and susceptibility: computes hazards
- Predictive hazards → Population clusters: clusters points
- Priority and allocation → Predictive hazards: uses risk tiers
- Evacuation planner → Shelter discovery: finds candidates
- Evacuation planner → Elevation-aware routes: requests routes
- Evacuation planner → Tactical maps: renders plans
- HTTP API → Flood agent: routes chat
- HTTP API → RAG report generator: generates reports
- HTTP API → Report knowledge store: queries context
- Flood agent → Agent tools: dispatches tools
- MirEye chatbot → MirEye physical data: queries context
- RAG report generator → Report knowledge store: retrieves reports
- Flood agent → LLM provider: consults model
- MirEye chatbot → LLM provider: consults model