FAQ — drawtonomy whiteboard for driving scenarios

Can I use drawtonomy to draw figures for an autonomous-driving paper, slide, or document?

Yes — it’s one of the main use cases. Lanes, intersections, vehicles, and pedestrians are built-in shapes, so a driving-scenario figure that would take half an hour in a generic drawing tool comes together in minutes. Export as drawtonomy.svg, PDF, or EPS for LaTeX, slide decks, and Markdown.

Is drawtonomy free to use?

Yes. drawtonomy.com runs entirely in your browser. There is no account, no install, and nothing is uploaded. The SDK and extensions are open-source on GitHub.

Is drawtonomy a free OpenSCENARIO and OpenDRIVE editor?

drawtonomy is a whiteboard that exports to ASAM OpenDRIVE 1.8 and OpenSCENARIO 1.3, including a single zip bundle ready for esmini playback. You sketch the scene visually — lanes, intersections, vehicles, paths — and drawtonomy generates the .xodr and .xosc files. See Export to OpenDRIVE / OpenSCENARIO / esmini.

Can drawtonomy edit Lanelet2 (.osm) maps in the browser?

Yes. drawtonomy can import Lanelet2 OSM files (including Autoware sample maps), edit lane geometry and connections visually, and export back to Lanelet2 OSM. It handles large maps, and for a quick edit you can import only the lanes you need. See Import a Lanelet2 (.osm) file.

Can I import a ROS occupancy grid map?

Yes. drawtonomy reads the ROS map_server format (.pgm + .yaml) used by nav2, Cartographer, Gmapping, and similar SLAM tools. The grid renders as a background layer so you can sketch lanes, paths, and obstacles directly on top. See Import a ROS OccupancyGrid.

How is drawtonomy different from generic drawing tools?

Generic drawing tools only know about generic shapes, so a lane is just a rectangle you redraw whenever the road turns. drawtonomy ships lanes, intersections, crosswalks, traffic lights, road markings, vehicles, and pedestrians as built-in shapes, and lanes carry Next / Previous / Left / Right connections. Drag a boundary and connected lanes follow. See Why drawtonomy and the lane connection model.

Do I need to install anything?

No. drawtonomy runs in any modern web browser. Open drawtonomy.com and start drawing. Extension developers can also use the local dev server (@drawtonomy/dev-server), which hosts the editor on localhost.

Can drawtonomy generate a driving scenario from natural language?

Yes. The AI Scene Generator extension reads a natural-language prompt or OpenSCENARIO XML and emits an editable canvas — lanes, vehicles, paths, and pedestrians as regular drawtonomy shapes. It supports Anthropic Claude, OpenAI GPT, and Google Gemini.

What file format keeps a drawtonomy scene re-editable?

The .drawtonomy.svg format — drawtonomy’s native SVG format. It is a regular SVG file — so it previews in browsers, GitHub, slide decks, and Markdown renderers — with metadata that drawtonomy uses to rebuild every shape, connection, shared point, and overlap relationship when you reopen the file. Saving as .drawtonomy.svg by default gives you both a previewable SVG and a re-editable source in one file. See Export your scene and the .drawtonomy.svg format reference.

Can I share a drawtonomy diagram with a link?

Yes. Open the menu, click Export, then Copy URL, and drawtonomy packs the whole scene — lanes, vehicles, connections, and the map origin — into a single link. Send it in chat, an email, or a GitHub issue, and whoever opens it sees the exact figure in their browser, fully editable. No account, no upload, and nothing is stored on a server — the diagram is encoded into the URL itself. Select shapes first to share only part of a scene. Very large scenes fall back to a same-browser local link; share those as a .drawtonomy.svg file instead. See Share a diagram by URL.

Can I add my own vehicle or sign templates?

Yes. Templates are SVGs registered in templates/manifest.json. Drop yours into the right folder (vehicle, pedestrian, road_marking, sign, other), open a PR, and it ships in the editor’s menu next to the built-in templates. See Contributing templates.

Does drawtonomy support large road networks?

Yes. drawtonomy handles large road networks, so you can import and edit sizeable maps. For a quick edit on a small section, importing only the lanes or roads you need keeps the editor snappiest.

Can I use drawtonomy diagrams in papers, slides, or documentation?

Yes. The drawtonomy.svg format is valid SVG, so it embeds in papers, slides, Markdown, and websites without conversion. PDF, PNG, JPG, and EPS are also available for traditional publishing pipelines.

Is drawtonomy open source?

The drawtonomy SDK, extension framework, templates, and this documentation site are open source on GitHub. The drawtonomy.com editor application is source-available, with the SDK and extensions as the OSS surface for building on top.

Does drawtonomy upload my drawing or scene data to a server?

No. drawtonomy runs entirely in your browser. The scene lives in browser memory and on your local file system when you save. Nothing is uploaded — no account, no telemetry on your shape data, no cloud sync. The AI Scene Generator extension sends prompts to the LLM provider you configure (Claude, GPT, or Gemini), but only the prompt text — not your existing scene.

Can I use drawtonomy offline?

Yes, after the first load. drawtonomy.com is a single-page application — once it has loaded in your browser, you can keep editing and save without a connection. Reopen the saved drawtonomy.svg later to continue. Satellite-background tiles and the AI Scene Generator extension are the only features that need the network.

Does drawtonomy support OpenSCENARIO 2.0 / DSL?

Not currently. drawtonomy exports OpenSCENARIO 1.3 XML, which is the version most simulators (CARLA, esmini, IPG CarMaker) consume today. OpenSCENARIO 2.0 / DSL is a different, parametric language and is on the roadmap but not yet implemented.

Can drawtonomy be used as an infinite canvas for driving scenarios?

Yes. drawtonomy is an infinite, pan-and-zoom canvas tuned for driving scenarios. Pan and zoom across arbitrarily large layouts, drop in satellite or roadmap backgrounds to trace from real locations, and the snap and shared-point system keeps lane geometry aligned without manual cleanup.

Can drawtonomy sketch ADAS scenarios like cut-in, lane change, or unprotected left?

Yes. These are routine sketches in drawtonomy. Cut-in, lane change, unprotected left turn, roundabout entry, pedestrian crossing, lead-vehicle braking, and similar functional / logical scenarios are sketched by dropping the road template, placing the ego and other actors, and adding path arrows for trajectories. See Sketching ADAS test scenarios for a worked example.

Where is the source code?

The SDK, extensions, templates, and this documentation site are on github.com/kosuke55/drawtonomy. The application itself is at drawtonomy.com.