amos-autonomous_mission_orchestration_system

AMOS — Autonomous Mission Orchestration System

“The LORD roars from Zion … the shepherd watches, and what he sees he declares.” — the Book of Amos

In the Bible, Amos was a shepherd called to be a watchman — scanning the horizon, seeing threats others missed, and speaking truth when it mattered most. That is exactly what this platform does: it watches across every domain, fuses what it sees, and orchestrates autonomous forces to protect and respond. The name isn’t a coincidence.

Autonomous Mission Orchestration System

AMOS is a multi-domain command-and-control platform that enables small teams of human operators to supervise and coordinate autonomous robotic assets across air, ground, maritime, cyber, and space domains.

Instead of controlling individual robots, operators define mission intent and AMOS orchestrates assets, sensors, autonomy engines, and communications to execute those missions.

v0.5.6 — Open Core Release | merkuri.one/amos

---

Run the Demo

Launch a fully simulated autonomous mission with one command:

git clone https://github.com/merkuriddg/amos-autonomous_mission_orchestration_system.git
cd amos-autonomous_mission_orchestration_system
./run_demo.sh

Browser opens automatically. Login: commander / amos_op1 — watch the mission unfold in real time.

Demo Scenarios:

• ./run_demo.sh border_patrol — Sensor network detects border intrusion, drones and ground robots investigate, track, and report (default)
• ./run_demo.sh swarm_recon — Drone swarm deploys into contested zone, adapts to jamming and battery loss, tracks high-value target
• ./run_demo.sh disaster_response — Earthquake SAR: drones map damage, robots inspect structures, AMOS prioritizes rescue tasks

Each scenario runs ~2 minutes with timed mission events, autonomous coordination, and human-in-the-loop checkpoints.

---

Platform Overview

Mission Console	Mission Planning	Simulation	Telemetry	Integrations
Threat Predictions	Kill Web	Swarm Control	SIGINT	Wargaming

---

Why AMOS Exists

Most robotics software focuses on controlling individual systems. AMOS focuses on coordinating autonomous teams.

• Multi-robot mission orchestration
• Autonomous swarm coordination
• Sensor fusion and autonomous cueing
• Human-machine teaming
• Resilient mesh communications
• Extensible plugin architecture

AMOS acts as the mission layer between robotics frameworks and operational applications — integrating mission planning, autonomy orchestration, sensor fusion, wargaming, and resilient networking into a single platform.

---

Quick Start

git clone https://github.com/merkuriddg/amos-autonomous_mission_orchestration_system.git
cd amos-autonomous_mission_orchestration_system
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
cp .env.example .env
python3 web/app.py

Open http://localhost:2600 — Login: commander / amos_op1

Optional: Database Setup

AMOS runs without a database (in-memory mode). For persistent storage:

bash db/setup.sh              # creates MariaDB database, user, schema, seed data

All Users

Username	Password	Role	Domain
commander	amos_op1	Commander	All
pilot	wings2026	Pilot	Air
grunt	hooah2026	Ground Op	Ground
sailor	anchor2026	Maritime Op	Maritime
observer	watch2026	Observer	All
field	tactical2026	Field Op	All

---

Editions

AMOS ships in two editions:

• Open (AMOS_EDITION=open) — Core C2 platform: 200+ API routes, map, assets, threats, EW, SIGINT, cyber, sensor fusion, mesh networking, voice commands, plugin system. Free and open-source.

• Enterprise (AMOS_EDITION=enterprise) — Full platform: 300+ API routes. Adds cognitive engine, NLP mission parser, Monte Carlo wargaming, swarm intelligence, kill web, ISR/ATR, effects chain, space domain, human-machine teaming, COMSEC, TAK/Link 16/STANAG integrations, and more.

See ENTERPRISE.md for the full comparison and licensing.

---

Architecture

Applications
  Mission Packs • Analytics • Operator Tools
        ▲
        │
AMOS Platform
  Mission Orchestration • Autonomy Engines • Swarm Coordination
        ▲
        │
Integration Layer
  ROS2 • MAVLink • TAK • MQTT • DDS • Link-16
        ▲
        │
Assets & Sensors
  Drones • Ground Robots • Maritime Vehicles • Satellites

amos/
├── web/                     — Flask app, routes, templates, simulation engine
├── core/                    — Data model, adapters, COMSEC, geo utilities
├── services/                — 48 autonomous subsystems
├── integrations/            — 23 protocol bridges (incl. DimOS bipedal robot bridge)
├── config/
│   ├── buildings/           — Building JSON floorplans (Alpha Compound, Embassy Bravo)
│   └── ...                  — Platoon config, theater locations, drone reference DB
├── examples/                — Demo scenarios + quickstart code samples
│   ├── border_intrusion/    — Border patrol demo
│   ├── swarm_recon/         — Swarm reconnaissance demo
│   ├── disaster_response/   — Earthquake SAR demo
│   └── quickstart/          — "Hello AMOS" API examples (Python + curl)
├── plugins/                 — 19 plugins (adapters, mission packs, examples)
│   ├── adsb_adapter/        — ADS-B aircraft surveillance
│   ├── aprs_adapter/        — APRS amateur radio position tracking
│   ├── atak_adapter/        — ATAK/TAK blue force & CoT
│   ├── ais_adapter/         — AIS maritime vessel tracking
|   ├── dragonos_adapter/    — DragonOS/WarDragon SDR sensor platform
|   ├── sdr_adapter/         — Unified SDR (SDR++ / SigDigger / DragonOS)
|   ├── remoteid_adapter/    — OpenDroneID / FAA RemoteID drone identification
|   ├── cot_adapter/         — Cursor-on-Target XML ingest (UDP multicast + TCP)
|   ├── zmeta_adapter/       — ZMeta v1.0 ISR metadata (UDP ingest + egress)
│   ├── meshtastic_adapter/
│   ├── px4_adapter/         — PX4 MAVLink autopilot bridge
│   ├── ros2_adapter/        — ROS 2 middleware bridge
│   ├── patrol_mission/      — Patrol mission pack (linear, orbit, sector sweep)
│   ├── search_rescue_mission/ — SAR mission pack (expanding square, parallel track)
│   ├── perimeter_security_mission/ — Perimeter defense (overwatch, tripwire zones)
│   ├── example_sensor/      — Reference sensor plugin
│   └── example_mission_pack/ — Reference mission template
├── sdk/                     — Developer SDK (Python package + docs)
│   ├── python/amos_sdk/     — Installable SDK (contracts, helpers, testing)
│   └── docs/                — Plugin tutorial, contracts ref, integration patterns
├── tools/                   — CLI tools (plugin scaffolding)
├── demo/                    — Legacy demo scenarios
├── db/                      — MariaDB schema + setup script (36 tables)
├── tests/                   — 537 automated tests
└── docs/                    — Architecture, SDK, simulation, API docs

---

Key Capabilities

Core Platform — 25 simulated assets (air/ground/maritime), 22 threats, C2 console with Leaflet map, digital twin dashboard, AWACS view, field 3D view, EW/SIGINT/cyber ops, swarm formation control, HAL autonomy engine, voice commands, contested environment simulation

Kill Chain & Readiness — F2T2EA kill web with human-approval gate, dynamic ROE posture, AI threat prediction, battle damage assessment, supply logistics

Wargaming — 1000+ iteration Monte Carlo simulations, Markov chain attrition modeling, 5 strategies × 3 tempos

Swarm Intelligence — Reynolds flocking physics, 6 emergent behaviors, task auction protocol, self-healing formations

ISR/ATR — Automatic target recognition, pattern-of-life analysis, change detection, prioritized collection

Effects Chain — Multi-domain strike orchestration (Cyber → EW → SIGINT → Kinetic → ISR), 5 pre-built templates

Space Domain — 9 orbital assets, Keplerian propagation, SATCOM link budgets, JADC2 mesh

Human-Machine Teaming — 5 autonomy levels, trust calibration, workload/fatigue assessment, adaptive delegation

Drone Reference DB — 105 drone entries (93 commercial via DroneCompare CC-BY-4.0, 7 military/tactical, 5 adversary), serial prefix lookup, RemoteID enrichment, counter-UAS context, searchable UI

CQB Operations Console — Full indoor mission planning and execution UI. Select buildings, generate clearing plans, assign assets, execute missions with real-time room clearance tracking. Threat intel overlay shows hostiles, IEDs, hostages, and civilian contacts from building intelligence and live perception fusion

3D Tactical Viewer — WebGL-rendered multi-story building viewer with floor isolation, color-coded room clearance status, asset position markers, threat icons, and SLAM occupancy grid overlay. Orbit/zoom/pan camera controls for full situational awareness

CQB Formations — 6 meter-scale tactical formations (STACK, BUTTONHOOK, CRISSCROSS, BOUNDING_OVERWATCH, PERIMETER, CORRIDOR) for close-quarters battle, dual lat/lng + local meter-coordinate modes, integrated with swarm orchestrator

Bipedal Squad Seeds — Environment type abstraction (outdoor_open, outdoor_urban, indoor_cqb), extended asset state model (posture, stance, manipulation, cover, fatigue, indoor position), indoor positioning data model for GPS-denied CQB ops

Building Data Model & Indoor Positioning — Per-structure JSON floorplans with rooms, doors, windows, stairwells, wall materials. Adjacency graph, BFS pathfinding, LOS queries, room clearing status. Multi-source indoor positioning (SLAM/UWB/IMU fusion) with per-asset tracking and lat/lng bridge for GPS-denied operations. Ships with 2 demo buildings: Alpha Compound (3-story) and Embassy Bravo (2-story diplomatic compound with threat intel)

CQB Room Clearing Planner — Autonomous mission planning for close-quarters battle: 6 CQB task types (BREACH, CLEAR, HOLD, SECURE, EXTRACT, STACK) with role assignment (point, number_2, cover, rear_security). Generates phased clearing plans from building models with dependency chains, reinforced door detection, and objective room targeting. CQB-specific ROE (hostage room restriction, fratricide prevention, CQB autonomy tier, range limits)

DimOS Bridge — Command-and-control integration for bipedal humanoid robots via the DimOS autonomy framework. Maps CQB task plans to robot navigation, manipulation, and perception primitives. Supports real robot connection (WebSocket) and standalone simulation mode. Configurable via DIMOS_ENDPOINT in .env. Manage from the Integrations page or CQB Ops console

System-of-Systems Dependency Map — Interactive D3.js force-directed graph showing every AMOS subsystem and how they connect. DB-backed nodes and edges with full CRUD. SWaP-C+ weighted scenario engine: 9 built-in weight dimensions (Risk, Reliability, Latency, Cost, Size, Weight, Power, Attrition, Geography) plus unlimited user-defined custom weights stored as JSON. Scenario planner lets operators save named weight-override sets, evaluate mission rollup scores (success probability, max risk, critical path latency, SWaP burden, Human Exposure Index, terrain sensitivity), and compare alternatives side-by-side. Red Team mode simulates node failures with cascading downstream impact visualization. Bush trail highlighting traces signal chains (CQB Assault, Sensor-to-Shooter, Resilience Backbone, Human Decision). Phase-band layout (Left of Bang → The Bang → Right of Bang) with node coloring by risk gradient in scenario mode, edge thickness by confidence, and drag-to-position persistence

Perception Fusion — Fuses detections from multiple indoor assets into a unified threat picture. Multi-sensor correlation (visual, thermal, LIDAR, acoustic), persistent track management, SLAM occupancy grid aggregation, and bidirectional intel forwarding to/from robots

Squad Supervisor — Autonomous squad mission orchestrator. Creates, plans, and executes supervised CQB missions with asset assignment, reserve management, formation control (STACK, LINE, WEDGE, DIAMOND, ECHELON, BOUNDING_OVERWATCH), contact handling, automatic re-tasking, and after-action report generation

Mesh Networking — MANET with 7 frequency bands, Dijkstra routing, frequency hopping, store-and-forward queuing

---

Documentation

Guide	Description
/api/docs	Interactive Swagger UI — auto-generated OpenAPI 3.0 docs for all routes
/api/v1/openapi.json	Machine-readable OpenAPI 3.0 JSON spec
/mobile	Mobile / Tablet Field Operator UI — touch-friendly responsive view
docs/platform/	System architecture, simulation guide, API versioning
docs/platform/AMOS_Plugin_SDK.md	Plugin SDK — build asset adapters and integrations
docs/platform/INTEGRATION_GUIDE.md	Integration bridge development
sdk/docs/PLUGIN_TUTORIAL.md	Step-by-step plugin building tutorial
sdk/docs/CONTRACTS.md	Data contract reference (events, types)
sdk/docs/INTEGRATION_PATTERNS.md	Bridge architecture and patterns
sdk/python/README.md	SDK quickstart and API reference
docs/hardware/DRAGONOS_SETUP.md	DragonOS / WarDragon SDR integration guide
docs/hardware/ZMETA_SETUP.md	ZMeta ISR metadata integration guide
ENTERPRISE.md	Enterprise edition features and licensing
SECURITY.md	Security policy and vulnerability disclosure
CONTRIBUTING.md	Contribution guidelines and code style
TRADEMARK.md	Trademark policy and usage guidelines

---

Tech Stack

• Backend: Python 3, Flask, Flask-SocketIO
• Frontend: Vanilla JS, Leaflet.js, CesiumJS, WebSocket
• Database: MariaDB (optional — runs in-memory without it)
• Testing: pytest (537 tests), GitHub Actions CI (Python 3.11 + 3.12)
• Integrations: MAVLink, CoT XML (send + receive), TADIL J, ROS 2, MQTT, DDS, Kafka, VMF, STANAG 4586, ADS-B, APRS, AIS, RemoteID, LoRa/Meshtastic, NMEA, DragonOS/WarDragon SDR, SDR++, SigDigger, ZMeta ISR Metadata, DimOS (bipedal robot bridge)
• Security: AES-256-GCM encryption, HMAC, key lifecycle management

Testing

python3 -m pytest tests/ -v --tb=short

537 tests across route, service, and contract layers. CI runs on every push via GitHub Actions.

Plugin Development

Build your own AMOS plugins with the scaffolding tool:

python3 tools/create_plugin.py my_sensor --type sensor_adapter

Supports 6 plugin types: asset_adapter, sensor_adapter, mission_pack, planner, analytics, transport

See the Plugin Tutorial for a step-by-step guide, or browse plugins/example_sensor/ for a reference implementation.

The SDK also provides typed data contracts and a test harness:

pip install -e sdk/python     # install AMOS SDK locally

from amos_sdk import PluginBase, SensorReading
from amos_sdk.testing import PluginTestHarness

See SDK README and Data Contracts for full reference.

---

License

AMOS Open Core is free and open source under the Apache License 2.0.

Enterprise modules are available under commercial license from Merkuri LLC. See ENTERPRISE.md.

Trademark

AMOS™ and Autonomous Mission Orchestration System™ are trademarks of Merkuri LLC. Use of these names in commercial products or services requires written permission. See TRADEMARK.md for the full policy.

All contributions grant Merkuri LLC rights to use contributions commercially.

Security

AMOS is a research and development platform. See SECURITY.md.

---

Get Involved

AMOS is built for the robotics, defense, and autonomous systems community. Here’s how to contribute:

• Build a plugin — Use tools/create_plugin.py to scaffold a new asset adapter, sensor, or mission pack
• Add an integration — Connect a new protocol or hardware platform (see integrations/ for examples)
• Improve docs — Tutorials, API examples, deployment guides
• Write tests — Expand coverage across route, service, and contract layers
• Try a demo — Run ./run_demo.sh and explore the platform

See CONTRIBUTING.md for guidelines.

---

Vision

Autonomous systems are rapidly evolving, but they remain fragmented. AMOS provides the mission operating system that allows autonomous systems to operate as coordinated teams.