Skip to content
drawtonomy drawtonomy

Equations next to autonomous-driving figures

This is the cross-over use case for drawtonomy’s two strongest features at once: the autonomous-driving shapes (lanes, vehicles, paths, intersections) and the Math (KaTeX) shape.

It comes up surprisingly often in self-driving research papers. You don’t just want a picture of the scenario. You want the picture and the equation that describes the model — the cost function, the controller, the kinematic update — so the reader can map symbols back to the world.

A typeset planning objective rendered with KaTeX on the same canvas as a driving scene

A KaTeX-rendered formula sits as a normal shape on the canvas — drop it next to lanes, vehicles, and trajectories to build a single self-contained figure.

When this matters

Section titled “When this matters”
  • Trajectory prediction. Show three predicted future paths for the ego vehicle, with the model’s loss function typeset next to them.
  • Motion planning. Show a candidate trajectory in the lane scene, with the planning objective \min \sum_t \| x_t - x_t^{ref} \|^2 + \lambda u_t^2 rendered alongside.
  • Control. Show a vehicle following a curve, with the feedback controller’s equation next to the relevant geometry.
  • Perception evaluation. Show a lane / pedestrian scene with the IoU or AP metric formula displayed next to it.
  • Behaviour models. A pedestrian-crossing scene with the social-force / IDM equation rendered next to the agent.

In each case, the equation is part of the figure, not a caption.

The drawtonomy workflow

Section titled “The drawtonomy workflow”

  1. Draw the scene first. Lanes with the Lane Tool, vehicles from the Vehicle templates, trajectories with the Path tool — exactly as in the Your first three lanes tutorial.

  2. Place the equation with the Math (fx) tool. Put it where it’ll read clearly — usually above or below the scene, or to one side.

    The fx (Math) button highlighted in the bottom toolbar

    The Math tool sits next to the Text tool in the bottom toolbar — / is the keyboard shortcut.

  3. Tie symbols to the scene. Use LineArrows or Text shapes to connect, say, the x_t in the equation to the relevant vehicle on the canvas. drawtonomy’s Snap feature helps the arrow latch to either side.

  4. Export PDF. \includegraphics{...} it into your paper. The math glyphs and lane/vehicle paths are all vectors in the same file.

Practical tips

Section titled “Practical tips”
  • Match the equation size to the scene size. A 20 px equation next to a 600 px lane diagram disappears. Bump the size slider to 32 – 48 px to read well at journal-figure scale.
  • Keep a .drawtonomy.svg source. When a reviewer asks to replace \sigma with \rho, double-click the equation and retype — the lanes and vehicles don’t move.
  • Equation numbering vs. caption numbering. If your figure has multiple equations and you want them numbered, use \begin{align} so the numbers (1)(2)(3) are part of the rendered SVG. Otherwise the paper’s \label{} system has no way to reference equations inside a figure.
  • Colour for emphasis. Recolouring the relevant term of the equation requires using two adjacent Math shapes (one black, one red) — drawtonomy renders one equation as a single shape, so per-token colour isn’t supported.

See also

Section titled “See also”