Skip to content
drawtonomy drawtonomy

Visualizing cut-in, lane-change, and unprotected-left scenarios

Every ADAS / AV test catalog has the same handful of recurring scenarios at its core — cut-in, lane change, unprotected left, roundabout entry, pedestrian crossing, lead-vehicle braking. The figure that goes into the test plan, the slide deck, the safety-case document, or the paper is almost always one of these, drawn from a top-down 2D view.

drawtonomy is reasonable for producing those figures quickly, with a consistent style across the whole catalog.

The recurring scenarios

Section titled “The recurring scenarios”

These are the functional scenarios that appear most often. Each is a short sketch once you have the right road template:

Cut-in scenario

Section titled “Cut-in scenario”

Another vehicle moves laterally from an adjacent lane into the ego lane in front of the ego vehicle. The figure shows:

  • A 2-lane (or 3-lane) road.
  • The ego in one lane, the cutting-in vehicle in the adjacent lane.
  • A path arrow from the cutting-in vehicle’s start to its end position in the ego lane.
  • The trigger label — typically TTC, relative speed, or longitudinal gap.

This is one of the most-tested ADAS scenarios because AEB (Automatic Emergency Braking) and ACC (Adaptive Cruise Control) systems are evaluated on it directly.

Lane change scenario

Section titled “Lane change scenario”

The ego vehicle changes lanes. The figure shows:

  • A multi-lane road.
  • The ego’s start position, end position, and trajectory.
  • Other actors that the manoeuvre interacts with (a lead vehicle being passed, a follower behind, an oncoming vehicle on the target lane in bidirectional cases).

Useful for evaluating lane-change planners, gap-acceptance models, and merge-assist systems.

Unprotected left turn

Section titled “Unprotected left turn”

The ego vehicle turns left across one or more lanes of oncoming traffic without a protected left-turn signal phase. The figure shows:

  • A 4-way intersection.
  • The ego’s left-turn trajectory.
  • One or more oncoming vehicles on the conflicting lane.
  • Optionally, pedestrians at the crosswalk on the ego’s destination leg.

A core scenario for urban AV planning research, and a frequent corner case in design reviews.

Pedestrian crossing scenario

Section titled “Pedestrian crossing scenario”

A pedestrian crosses the road in front of the ego, with varying visibility and timing. The figure shows:

  • A road segment with a marked or unmarked crossing.
  • The pedestrian’s trajectory.
  • The ego’s approach.
  • Optionally, an occluding vehicle that hides the pedestrian until late.

Pedestrian crossings are at the heart of perception, prediction, and Vulnerable-Road-User-protection arguments.

Lead-vehicle braking

Section titled “Lead-vehicle braking”

A vehicle in front of the ego brakes. Used to evaluate following gap, brake-light detection, and AEB. The figure shows:

  • A single lane.
  • The ego and the lead vehicle.
  • A braking annotation on the lead vehicle.
  • The ego’s expected response.

Roundabout entry

Section titled “Roundabout entry”

The ego enters a roundabout with circulating traffic. Used to evaluate yield logic, gap acceptance in non-Manhattan geometry, and lane-curvature handling. The figure shows:

  • The roundabout geometry.
  • The ego’s approach lane and exit lane.
  • One or more circulating vehicles.
  • The yield line and any pedestrian crossings on the approach / exit.

A reusable starter set

Section titled “A reusable starter set”

For each of the recurring scenarios, build the road template once and save it as a .drawtonomy.svg file. drawtonomy preserves lane topology in the .drawtonomy.svg metadata, so the geometry stays correct under future edits. The template files end up as a small library of reusable scenes:

  • 2-lane-highway.drawtonomy.svg
  • 3-lane-highway.drawtonomy.svg
  • 4-way-unprotected.drawtonomy.svg
  • roundabout.drawtonomy.svg
  • t-junction.drawtonomy.svg
  • urban-arterial-with-crosswalk.drawtonomy.svg

When a scenario variant is needed (different parameter values, different lane count), open the matching .drawtonomy.svg template, drop in the actors at the new positions, and re-export.

Style suggestions

Section titled “Style suggestions”

A few things that help across a catalog of figures:

  • Consistent ego style. Pick one colour for the ego and stick with it across the whole catalog. Readers learn to find it at a glance.
  • Direction-of-travel arrows. Use Path arrows pointing in the direction of motion. Avoid bidirectional arrows unless the scenario genuinely has bidirectional motion.
  • Short trigger labels. “TTC = 2.5 s” beats a sentence. Put the explanation in the surrounding text, not on the figure.
  • Grayscale-safe palette. Many journals still print in grayscale. The Attribute Panel lets you separate colour from opacity / stroke so the figure stays readable when colour is removed.

What this isn’t for

Section titled “What this isn’t for”
  • Parameter sweeps. Each variant is a separate sketch. If you need 100 parameter combinations, generate them from a DSL or library like scenariogeneration and only sketch the canonical figure here.
  • Executable scenarios. drawtonomy’s OpenSCENARIO 1.3 export covers a subset of the spec (per the exporter docs) — no parameter sweeps, no conditional triggers, no complex storyboards. For executable test scenarios that go into a regression suite, hand-edit XML or generate from a DSL.
  • Photorealistic rendering. drawtonomy is strictly top-down 2D. Use a simulator screenshot for that.
Section titled “Related reading”