Feature: FET_036_RTX#

Property

Value

Feature Name

FET_036_RTX

Runtime

ISAAC

Proprietary Techs

None

Latest Version

0.1.0

Description#

Defines the simulation-readiness contract for OmniLidar RTX sensor prims. LiDAR is an RTX rendering sensor that fires rays and produces point cloud data via the GPU render pipeline. It is categorized under nonvisual_sensors/ because its output is rendered data (GenericModelOutput point cloud), not physics simulation state.

Dependency Graph#

This feature has no dependencies and no other features depend on it directly.

Use Cases#

Products or workflows that consume this feature:

  • SimReady validation verifies LiDAR sensor schema and parameter consistency at asset authoring time.

  • Isaac Sim runtime benchmark test (lidar_point_cloud) verifies the sensor produces accurate range measurements by surrounding it with cubes at known distances and checking that measured returns match expected wall distances.

Requirements#

Version 0.1.0#

Details

Used in Profiles#

  • Sensor-LiDAR v1.0.0 (required)

Feature Dependencies#

None.

Requirement List#

Requirement

Requirement Doc

Rule

LI.001

LI.001

Implementation

LI.002

LI.002

Implementation

LI.003

LI.003

Implementation

LI.004

LI.004

Implementation

Pipelines#

Source file type:

  • .usd, .usda, .usdc

Validation or runtime pipeline:

  • simready-validate verifies FET_036_RTX requirements at asset authoring time.

  • simready-benchmark --features FET_036_RTX runs RTX LiDAR point cloud verification in Isaac Sim (requires GPU driver >= 576.x).

Samples#

Benchmarks#

  • lidar_point_cloud — surrounds each OmniLidar sensor with 16 cubes at asymmetric known distances, runs 120 RTX scan frames, and verifies that per-return measured ranges match expected wall distances. Passes when fewer than 1% of non-edge returns exceed 2% range error. Skips (not fails) when GPU driver does not support RTX LiDAR CUDA 12.9.

Adapters#

None.