Python: Radar Sensor#

This example loads radar_example.usda, renders a radar PointCloud, maps the composite render variable to CPU memory, and visualizes detections in rerun.io. Point color is derived from signed RadialVelocityMs: blue is approaching the sensor, green is near zero, and red is receding.

The scene is Z-up and contains a radar at (0, 0, 1) rotated to look along world +X, an asphalt ground plane, a moving steel cube, and a fixed concrete cube. The moving cube advances toward the radar across 10 simulation steps.

“Create a Python sensor example that loads a scene containing a configured radar and moving target, advances scene time over multiple steps, reads valid detections and signed radial velocity, prints per-step summaries, and optionally visualizes detections with velocity-based colors.”

Radar sensor example output

Prerequisites#

  • Python 3.10-3.13

  • uv

Running#

cd examples/python/sensors/radar
uv run main.py

Options#

Flag

Description

--scene PATH

Load a different USDA scene

--steps N

Number of animated radar steps

--no-rr

Disable Rerun visualization and print only the per-step summary

--rrd PATH

Write a Rerun recording instead of spawning a viewer

Expected console output values vary, but a successful run prints per-step valid point counts and radial velocity min/max values.