What is this?
WWW-POSIM is a simulator for exploring marine environments and testing how marine robots move. It builds coastal and seabed terrain from geographic data, lets you operate surface vehicles such as WAM-V and underwater vehicles such as BlueROV2, and shows their sensor data. Use it to practise operating a robot or test a controller before working with physical equipment.
You can use it on the website or install the app on your computer. A world is a virtual test environment containing terrain, water and robots.
See Specifications for default sensors, ROS messages, terrain sizes and software versions.
WWW-POSIM stands for World Wide Web Platform for Ocean Simulation. WWW means World Wide Web, and POSIM means Platform for Ocean Simulation.
The name reflects the long-term goal of making the platform's full capabilities available as a web service. The current web service has performance and resource limits, so installed apps are also available. For demanding simulations and higher-quality graphics, we recommend the installed app, which uses your computer's processing power and supported GPU.
What can you do?
| Feature | Use it to | Guide |
|---|---|---|
| Coastal and seabed terrain | Build a test area from latitude and longitude | Generate terrain for your site |
| Surface and underwater robots | Operate WAM-V, dive with BlueROV2 and follow waypoint routes | Workspace basics |
| Sensors and control | Observe cameras, LiDAR and IMU data; test Python or ROS 2 controllers | Control an installed workspace |
| Custom robots | Add models and configure mass, thrusters and sensors | Prepare your vehicle |
| Public LIVE | Watch the lead and following vehicles and their sensor feeds | LIVE world voyage |
Website or installed app?
We recommend the installed app for your first simulation. It uses your computer's processing power and supported GPU, with higher graphics quality than the website.
| Mode | Where the simulation runs | How to start |
|---|---|---|
| Installed app | Your computer | Install the app and sign in |
| Personal workspace on the website | Server | Sign in on the website and start a simulation |
| Public LIVE viewing | Shared server simulation | Choose LIVE on the main page |
A workspace is the screen where you create and run worlds. The website can run only a limited number of simulations at once, so you may need to wait when it is busy. The installed app and personal web workspaces require an account with a verified email address. See Installation for operating-system support and Account and usage time for time allowances.
How POSIM fits in
POSIM is the simulation library providing marine environments, robot models and sensors.
WWW-POSIM keeps the upstream POSIM source unchanged as a Git submodule, pinned to a selected revision. Its application code and separately maintained libraries add:
Current WWW-POSIM capabilities
- Access POSIM through a browser workbench or installed applications on Windows, macOS and Linux, with isolated personal simulation sessions.
- Generate geographic worlds from elevation, bathymetry, coastline and imagery data, with terrain prepared along a route and cached for reuse.
- Explore vehicles and worlds through browser-rendered or Gazebo-rendered views, with maps, instruments, ground controls, planned routes and recorded tracks.
- Configure WAM-V and BlueROV2 vehicles, set world and vehicle starting positions independently, and save or import vehicle configurations and models.
- Customize vehicle mass, inertia, buoyancy, hydrodynamics, thrusters and supported sensors.
- Inspect camera, LiDAR, IMU, GNSS, pressure, thrust and other sensor data. DVL and sonar options depend on the required providers being installed.
- Run ArduPilot SITL missions, including waypoint navigation, station keeping, inspection, survey and dive-and-return, with mission controls and terrain-based route checks.
- Connect QGroundControl to personal Autopilot sessions for vehicle control, parameter access and selected-camera video, with installation and launch support. The installed app must remain running for both local and web connections. QGroundControl
- Control vehicles with an on-screen joystick or a USB/Bluetooth gamepad, with adjustable output and safeguards for interrupted input.
- Connect custom controllers and sensor tools through the Python SDK and ROS 2, and use native RViz2, browser-based RViz and Foxglove viewers.
- Configure waves, currents, depth profiles, simulation speed and visual settings, and inspect performance information.
- Export world SDF files and re-import supported settings; manage saved terrain and vehicle revisions with archive and restore.
- Follow public LIVE voyages through rendered views, sensor feeds and route history, with AIS traffic where available. AIS vessels are visual only and have no collision response.
- Learn and test autonomous navigation with six teaching tasks, downloadable examples, automatic scoring, result playback and public-alias leaderboards. WWW-POSIM autonomous navigation challenges
- Use multilingual interfaces, guides, technical references and documentation search.
Route checks cover terrain and hull clearance; they do not provide full moving-vessel avoidance or COLREG compliance. AIS ships are visual-only and depend on source coverage. Optional sensors depend on the installed provider. See Supported features for session differences.
In development, not yet released or fully validated
- Multi-vehicle worlds with independent namespaces, ports, sensors and controls, plus saved fleet layouts.
- USV–UUV and USV–UAV cooperation with Tethys/LRAUV and Iris, including following and WAM-V formations.
- Moving-deck UAV operations and RobotX-inspired practice-arena assets, with full end-to-end acceptance still in progress.
- Visual, inertial and acoustic SLAM workflows, map inspection and export, and VRX practice worlds, with platform and sensor validation still in progress.
- Importable simulation workflows and consolidated results. Automatic native execution, sensor recording and action replay are still being developed.
- Companion-computer and backseat-driver workflows with SITL. Physical Raspberry Pi/Jetson setups and FCU hardware-in-the-loop operation have not yet been validated.
- Comparisons against real sonar and vehicle datasets. This work does not yet establish the simulator’s real-world accuracy.
Gazebo computes vehicle physics and sensor data. You can view a Gazebo-rendered scene or an interactive browser view. Underwater vehicles use buoyancy and hydrodynamic models; surface WAM-V uses buoyancy, drag and wave models. See Components & libraries and Custom worlds for details.
Read next
Choose English, Spanish, French, German, Italian, Korean, Japanese or Chinese from the language selector. Documentation search covers titles, body text, tables, commands and parameters. Each result opens the matching section.