Known Issues

Known Issues#

Attention

Please also refer to the Omniverse Isaac Sim documentation for known issues and workarounds.

Each entry below names the backends it affects. An issue listed under one backend does not apply to the others unless it says so.

PhysX backends#

Surface grippers require CPU simulation#

Affects: physics=isaacsim_physx surface-gripper tasks.

Surface grippers require CPU simulation. This includes the UR10 Long/Short Suction stacking tasks, the Galbot Right Arm Suction stacking task, and its relative and absolute Mimic variants. Pass --device cpu when running teleoperation. Zero and random agents preserve these tasks’ CPU defaults when --device is omitted; an explicit GPU override is unsupported.

Sensor readings are stale immediately after a reset#

Affects: physics=isaacsim_physx, physics=ovphysx, and any RTX-based renderer.

Many physics engines do a simulation step as a two-level call: forward() and simulate(), where the kinematic and dynamic states are updated respectively. PhysX has only a single step() call where the two operations are combined. Because of computations through GPU kernels, it is not straightforward to split them. As a result, writing a root or joint state does not by itself run a full forward pass.

For articulation link poses this is handled: reading body_link_pose_w (or the deprecated body_state_w) triggers a PhysX kinematic update, so link poses reflect a preceding root-state or joint-state write without an intervening step().

For RTX rendering-based sensors — cameras in particular — the data is still not refreshed by a state write. The rendering engine update is bundled with the simulator’s step() call, so the sensor data is only refreshed when the simulation is stepped forward, and a read taken between a reset and the next step returns the previous frame.

There is currently no direct workaround for the sensor case. From our experience, the reset values affect agent learning in proportion to how frequently the environment terminates; as an agent learns successfully, the termination rate drops and the effect becomes less significant.

Exiting the process#

Affects: physics=isaacsim_physx.

When exiting a process with Ctrl+C, occasionally the below error may appear:

[Error] [omni.physx.plugin] Subscription cannot be changed during the event call.

This is due to the termination occurring in the middle of a physics event call and should not affect the functionality of Isaac Lab. It is safe to ignore the error message and continue with terminating the process. On Windows systems, please use Ctrl+Break or Ctrl+fn+B to terminate the process.

Newton backends#

OpenUSD can crash while parsing collider-rich rigid bodies#

Affects: Kitless Newton workflows using OpenUSD releases earlier than 26.05, and Isaac Sim releases earlier than 6.1. Isaac Sim 6.1 and later contain the upstream fix in their bundled OpenUSD runtime.

Older OpenUSD releases have a thread-safety issue in UsdPhysics.LoadUsdPhysicsFromRange when multiple colliders belong to the same rigid body. Newton uses this API while importing a USD stage, so affected processes can terminate during scene creation with a segmentation fault, access violation, heap-corruption message, or hang. The issue and upstream fix are described in OpenUSD PR #4002.

Note

For kitless workflows, Isaac Lab obtains the pxr modules from usd-exchange. usd-exchange and usd-core are alternative Python distributions of the same OpenUSD runtime and should not be installed together because both provide pxr. They use different distribution version schemes: for example, usd-exchange==2.3.0 provides OpenUSD 25.05. Isaac Sim instead uses its own Kit-bundled OpenUSD runtime.

As a workaround for an affected kitless or pre-6.1 Isaac Sim runtime, limit OpenUSD to one worker thread. Set PXR_WORK_THREAD_LIMIT before launching Python so the limit is present before any pxr module initializes:

PXR_WORK_THREAD_LIMIT=1 uv run isaaclab train --rl_library rsl_rl --task Isaac-Reach-Franka physics=newton_mjwarp

set PXR_WORK_THREAD_LIMIT=1
uv run isaaclab train --rl_library rsl_rl --task Isaac-Reach-Franka physics=newton_mjwarp

This environment variable limits OpenUSD’s process-wide worker pool to one thread and can reduce USD import performance. Use it only with affected runtimes, and remove it after upgrading to an OpenUSD provider that contains the fix or to Isaac Sim 6.1 or later.

Closed-loop articulations are not available on Newton (e.g. Agility Digit)#

Affects: physics=newton_mjwarp, physics=newton_kamino.

Robots whose USD encodes a closed kinematic loop — such as the achilles rod and toe push-rods on the Agility Digit — are not currently validated on the Newton backends. The Digit-based contrib tasks are PhysX-only and do not expose a Newton preset at all:

  • IsaacContrib-Velocity-Flat-Digit

  • IsaacContrib-Velocity-Rough-Digit

  • IsaacContrib-Tracking-LocoManip-Digit

Passing presets=newton_mjwarp to these tasks is rejected, because the preset never selected a Newton backend on them; it only stripped a center-of-mass randomization from a PhysX run. Use the default PhysX configuration for Digit-based environments.

Renderers#

Blank initial frames from the camera#

Affects: RTX-based renderers (renderer=isaacsim_rtx, renderer=ovrtx).

When using the Camera sensor in standalone scripts, the first few frames may be blank. This is a known issue with the simulator where it needs a few steps to load the material textures properly and fill up the render targets. It is most likely on a cold asset cache and in scenes with many or large textures; simple scenes with locally cached assets often render content on the very first frame.

If you do see blank frames, add the following after initializing the camera sensor and setting its pose:

from isaaclab.sim import SimulationContext

sim = SimulationContext.instance()

# note: the number of steps might vary depending on how complicated the scene is.
for _ in range(12):
    sim.render()

Animated curves disappear under Isaac RTX scene partitioning#

Affects: renderer=isaacsim_rtx with scene partitioning enabled. The OVRTX backend updates animated-curve bounding boxes correctly and is unaffected.

Cables are authored as UsdGeom.BasisCurves and animated every frame. Kit RTX computes a bounding box for a curve prim once and never refreshes it while the curve deforms (OMPE-105749). The per-environment scene partition is sized from the union of the bounding boxes of the prims it contains, so that union covers the cable’s initial extent plus whatever static geometry shares the partition. Once the cable moves outside that union it is culled and vanishes from the tiled camera images — for example a cable resting on a table can disappear the moment a robot arm pushes it off the edge.

The bug is specific to the Isaac RTX (Kit) backend with enable_scene_partitioning enabled.

The displacement needed to trip the cull is the distance from the curve’s spawn extent to the edge of its partition’s bounding box, so it depends on what else shares the partition. In a measurement on Kit 110.1.2, the cable in Isaac-Lift-Cable-Franka-Camera vanished at 0.6 m of displacement, while a lone curve in a partition containing only itself and a camera survived to roughly 4 m. Smaller motions render normally, which is why a settled or lightly perturbed cable looks fine.

There are two workarounds:

  • Pin the partition bounds. Spawn a pair of millimetre-scale static cubes at diagonally opposite corners of a box that conservatively envelops the environment’s workspace. They enlarge the partition’s bounding-box union to that box, so the cable stays inside it wherever it moves. Isaac-Lift-Cable-Franka and Isaac-Lift-Cable-Franka-Camera ship this workaround as the partition_bounds_marker_min and partition_bounds_marker_max scene entries; copy the pattern into custom cable environments. With the markers in place, cable visibility matches an unpartitioned render exactly. The markers are static visual prims without colliders, so they do not participate in physics, and Isaac-Lift-Cable-Franka-Camera drops them when its camera renderer has scene partitioning off.

  • Disable scene partitioning. Set enable_scene_partitioning to False to opt out of partitioning entirely, at the cost of the per-environment culling.

Scene partitioning is capped at 15625 partitions#

Affects: renderer=isaacsim_rtx with scene partitioning enabled, and renderer=ovrtx.

The underlying rtx.scenedb.plugin allocates a fixed-size pool of scene partitions and caps it at 15625, regardless of which renderer requests them. Isaac Lab assigns one scene partition per environment when enable_scene_partitioning is enabled for the Isaac RTX renderer, and OVRTX always assigns one scene partition per environment, so runs with more than 15625 environments exceed the pool on either backend. Once the cap is hit, rtx.scenedb.plugin logs a warning and discards the remaining partitions:

[Warning] [rtx.scenedb.plugin] SceneDbContext : Maximum number of scene partitions
(15625) reached. Additional scene partitions will be discarded.

Environments beyond the cap are left without their own partition and end up sharing one with another environment, so their tiled camera views can render another environment’s geometry instead of their own.

Using instanceable assets for markers#

Affects: all Kit-based renderers.

When using instanceable assets for markers, the markers do not work properly, since Omniverse does not support instanceable assets when using the UsdGeom.PointInstancer schema. This is a known issue and will hopefully be fixed in a future release.

If you use an instanceable assets for markers, the marker class removes all the physics properties of the asset. This is then replicated across other references of the same asset since physics properties of instanceable assets are stored in the instanceable asset’s USD file and not in its stage reference’s USD file.

Asset import#

URDF importer: unresolved references for fixed joints#

Affects: the URDF importer, independent of the physics backend.

Links connected through fixed_joint elements are not merged when their URDF link entries specify mass and inertia, even if merge-joint is set to True. This is expected behaviour — those links are treated as full bodies rather than zero-mass reference frames. However, the USD importer currently raises ReportError warnings showing unresolved references for such links when they lack visuals or colliders. This is a known bug in the importer; it creates references to visuals that do not exist. The warnings can be safely ignored until the importer is updated.

Environment and setup#

GLIBCXX errors in conda environments#

Affects: conda-based installations, independent of the physics backend.

Some workflows exit with an OSError indicating version 'GLIBCXX_3.4.30' not found when running from a conda environment. The issue appears to stem from importing torch or torch-related packages, such as tensorboard, prior to launching AppLauncher. As a workaround, ensure that all torch imports happen after the AppLauncher instance has been created, which should resolve the error.