Quickstart
Create a Project from a template, ask the Agent to change the Scene, run a radar simulation and inspect the Result.
This walkthrough takes about five minutes. It uses the Radar Sim template, which needs the Radar Plugin and an NVIDIA GPU. With no GPU, use the Empty Scene template and skip the solver step.
Create a Project from a template
With no Project open, Studio shows the chat and a grid of template cards. Click Radar Sim. Studio creates the Project in
Documents/WiTwin Projects/Radar Sim, opens it and starts a Session on its main Scene.A card with a download mark needs a Plugin that is not installed yet. Clicking it opens the Plugins page so you can install it first.
Look around
The viewport shows a 77 GHz FMCW radar facing one box target. The Hierarchy lists the Scene's objects; select one to edit its components in Properties. The left column lists your Projects, and each Project's Sessions under it.
Ask the Agent for a change
In the chat, type for example:
Move the box 2 m further from the radar and add a second box 1 m to its left.The Agent edits the live Scene. Each edit appears in the viewport as it happens and can be undone with Ctrl+Z.
Run the simulation
Ask the Agent to run the radar simulation, or start the Radar solver from its panel. Progress shows while it runs, and the finished run appears in the Results panel with its plots and data.
Commit your work
The Project is a git repository. Open the Git view from the top bar to see the changed Scene file, review the object-level diff and commit.
What you created#
Radar Sim/
witwin.toml # Studio manifest: name, main scene
pyproject.toml # Python project; lists the Radar Plugin as a dependency
README.md
scenes/main.wtscene # the Scene, as text
scripts/ # scripts that use the Scene without Studio
tests/ # pytest checks of the Scene
.witwin/ # local runtime state, git-ignoredThe template's scripts use the public witwin.studio package, so the same
simulation runs from a terminal. See Run without the editor.
Next steps#
- The Agent and Sessions: what the Agent can do and how it asks before risky actions.
- Parameters and formulas: drive dimensions from named values.
- Design optimization: tune those values automatically.