How nicholasaleks/damn-vulnerable-drone works
Project Type and Purpose
• This project is a full-scale drone hacking simulator built around the ArduPilot/MAVLink environment. It intentionally introduces vulnerabilities to provide a simulated environment where offensive security professionals and pentesters can practice drone attack techniques.
• The README clearly describes the goal of emulating various drone components (flight controller, companion computer, ground control station, simulator via Gazebo, etc.) in Docker containers. The system also supports different deployment modes (Full-Deploy vs. Half-Baked).
Analyzing the File Structure
• There are several top-level directories:
• “companion-computer”: Contains the web interface, configuration files (e.g., DHCP, DNS, SSH, MAVLink router), endpoints (routes), static and template resources.
• “flight-controller”: Contains the ArduPilot configuration and scripts to initialize the flight controller.
• “ground-control-station”: Contains configuration (e.g., wpa_supplicant) and mission/stage scripts.
• “qgc”: Contains launcher scripts and configs for QGroundControl.
• “simulator”: Has Docker configurations, Gazebo assets (3D models, worlds) as well as a management interface (mgmt folder) and a gzweb subcomponent.
• The presence of Dockerfiles in several directories and a top-level docker-compose.yaml indicates that each major component is containerized.
• The “simulator/mgmt” folder further confirms that there is a dedicated management web console for initializing flight states and triggering attack scenarios.
• Flight Controller: Represents the ArduPilot firmware running in a container. Mark it with its Docker IP (10.13.0.2) with connection lines to the simulator.
• Companion Computer: Handles higher-level processing like telemetry, camera streaming, and interfaces with both the flight controller and the ground control station. It is assigned Docker IP 10.13.0.3 and a simulated wireless IP (192.168.13.1).
• Ground Control Station (GCS): Provides the remote pilot’s interface, mission planning and control. It has Docker IP 10.13.0.4 and wireless IP (192.168.13.14). Indicate that it communicates via the MAVLink protocol.
• Simulator (Gazebo): Represents the 3D simulation environment and the underlying physics engine. It communicates with the flight controller and is usually running at Docker IP 10.13.0.5.
• Management Web Console: Although not explicitly a separate container by name, it is part of the “simulator/mgmt” folder, and provides UI control to trigger flight states and attack scenarios.
• Additional components: QGroundControl integration and supporting Docker overlays (e.g., for the “qgc” container) can be included for completeness.
• Simulated networks: the dedicated simulation network (10.13.0.0/24) as well as the simulated wireless network (192.168.13.0/24) used between the companion computer and the ground control station.
Architectural Patterns and Design Principles
• Microservice or containerized architecture – each major component runs in its own Docker container.
• That the system follows a layered approach where lower-level systems (flight controller) interface with higher-level ones (companion computer) and then the UI layer (ground control and management web console).
Happy diagramming!