How shashank101-aiml/maritime-control-tower works
Maritime Agentic Control System is a governed maritime logistics control tower. Operators use the React console to monitor vessels and events, analyze corridor risk, compare routes, and review recommendations; the backend workflow coordinates agents from ingestion through risk and routing to decisions and explanations. Human review, authorization, execution traces, and audit are governance capabilities alongside the main workflow. Separate capabilities cover fleet tracking, congestion, delay, fuel, and disruption simulation. The README documents these responsibilities; sampled source excerpts are partial, so the graph omits uncertain implementation-level links.
Main entry points: maritime_workflow.py Workflow coordinator · ingestion_agent.py Live ingestion · fleet_monitor_agent.py Fleet monitoring · core.py Agent governance
Components 27
Operator console
- React console
frontend/src/App.jsx - Route planning
frontend/src/pages/RouteRecommendations.jsx - Route service
frontend/src/services/routeService.js
API and workflow
- FastAPI app
backend/app/main.py - Workflow endpoint
backend/app/api/routes/workflow.py - Workflow coordinator
backend/app/workflows/maritime_workflow.py
Maritime analysis
- Live ingestion
backend/app/agents/ingestion/ingestion_agent.py - Sea conditions
backend/app/agents/ingestion/live_conditions_client.py - AIS positions
backend/app/agents/ingestion/ais_client.py - Risk scoring
backend/app/agents/risk/risk_agent.py - Route optimization
backend/app/agents/route/route_agent.py - Recommendation
backend/app/agents/decision/decision_agent.py - Explanation
backend/app/agents/explanation/explanation_agent.py - News understanding
backend/app/agents/understanding/event_understanding_agent.py - What-if simulation
backend/app/agents/simulation/simulation_agent.py
Fleet and estimates
- Port congestion
backend/app/agents/congestion/congestion_agent.py - Shipment delay
backend/app/agents/delay/delay_agent.py - Fuel and emissions
backend/app/agents/fuel/fuel_agent.py - Fleet monitoring
backend/app/agents/fleet/fleet_monitor_agent.py - Digital twin
backend/app/twin/digital_twin.py
State and governance
- Agent governance
backend/app/governance/core.py - Audit and trace
backend/app/governance/audit.py - Application database
backend/app/database/session.py
Other components
- Operator
- Weather feed
- AIS feed
- Human reviewer
Connections 28
- Operator → React console: uses
- React console → Route planning: opens planning
- Route planning → Route service: requests options
- Route service → FastAPI app: calls API
- FastAPI app → Workflow endpoint: dispatches routes
- Workflow endpoint → Workflow coordinator: runs workflow
- Workflow coordinator → Live ingestion: collects inputs
- Live ingestion → Sea conditions: fetches conditions
- Sea conditions → Weather feed: reads feed
- Workflow coordinator → Risk scoring: scores hazards
- Workflow coordinator → Route optimization: ranks routes
- Workflow coordinator → Recommendation: forms recommendation
- Workflow coordinator → Explanation: explains result
- Live ingestion → AIS positions: collects positions
- AIS positions → AIS feed: reads positions
- Operator → News understanding: submits news
- Operator → What-if simulation: requests scenario
- Operator → Port congestion: requests analysis
- Operator → Shipment delay: requests estimate
- Operator → Fuel and emissions: requests estimate
- Operator → Fleet monitoring: registers vessels
- Fleet monitoring → AIS positions: tracks ships
- Route optimization → Digital twin: routes over
- Operator → Human reviewer: reviews action
- Human reviewer → Agent governance: approves or rejects
- Agent governance → Audit and trace: records execution
- Agent governance → Application database: persists records
- Fleet monitoring → Application database: stores fleet data