Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Randomization, noise and delay profiles

Shipped task configs use these randomization, delay, and noise settings:

Task/configRandomizationDelayNoiseProvenance
waypoint_navigationtarget_pose_evolution, a tagged task event at a 0.05 s intervalAction (0, 3) and observation (0, 1) agent steps; per-reset draws use env.seedWaypoint-relative observation noise: 0.01 m / 2.5° episode bias plus 0.0025 m / 0.5° per-step noiseLegacy task constants; uncalibrated
locomotion_waypoint_navigationInherits target_pose_evolutionInherits action (0, 3) and observation (0, 1) agent stepsInherits waypoint-relative observation noise: 0.01 m / 2.5° episode bias plus 0.0025 m / 0.5° per-step noiseInherited waypoint task config; uncalibrated
waypoint_navigation with env.sim_to_real_profile=waypoint_navigation_synthetic_v1Same target eventAction and observation (1, 2) agent steps, fixed for each episode; 25 Hz means 40–80 msGaussian action noise, standard deviation 0.02 in normalized policy-action units; the waypoint observation noise above remains in metres and radianssynthetic_unmeasured
Other shipped task configsMany define tagged or untagged event termsNo other task config currently enables nonzero action or observation delayNo default task config enables Isaac Lab action_noise_model or observation_noise_modelPer-task terms remain their existing contracts

The opt-in waypoint fixture can be selected with:

srb agent train --algo sbx_ppo --env waypoint_navigation \
  env.sim_to_real_profile=waypoint_navigation_synthetic_v1 \
  env.seed=20260908

Delay ranges are integer agent steps. At the profile’s 25 Hz rate, one step is 0.04 s. The profile requires agent_rate=0.04 and rejects a custom action_noise_model.

This profile is a synthetic fixture, not hardware calibration or evidence of sim-to-real readiness. Calibration requires measured latency and error data.