Merge branch 'main' into patch-1

README.md

@@ -43,6 +43,19 @@ We wholeheartedly welcome contributions from the community to make this framewor
 These may happen as bug reports, feature requests, or code contributions. For details, please check our
 [contribution guidelines](https://isaac-sim.github.io/IsaacLab/main/source/refs/contributing.html).
 
+## Show & Tell: Share Your Inspiration
+
+We encourage you to utilize our [Show & Tell](https://github.com/isaac-sim/IsaacLab/discussions/categories/show-and-tell) area in the
+`Discussions` section of this repository. This space is designed for you to:
+
+* Share the tutorials you've created
+* Showcase your learning content
+* Present exciting projects you've developed
+
+By sharing your work, you'll inspire others and contribute to the collective knowledge
+of our community. Your contributions can spark new ideas and collaborations, fostering
+innovation in robotics and simulation.
+
 ## Troubleshooting
 
 Please see the [troubleshooting](https://isaac-sim.github.io/IsaacLab/main/source/refs/troubleshooting.html) section for
@@ -56,6 +69,15 @@ or opening a question on its [forums](https://forums.developer.nvidia.com/c/agx-
 * Please use GitHub [Discussions](https://github.com/isaac-sim/IsaacLab/discussions) for discussing ideas, asking questions, and requests for new features.
 * Github [Issues](https://github.com/isaac-sim/IsaacLab/issues) should only be used to track executable pieces of work with a definite scope and a clear deliverable. These can be fixing bugs, documentation issues, new features, or general updates.
 
+## Connect with the NVIDIA Omniverse Community
+
+Have a project or resource you'd like to share more widely? We'd love to hear from you! Reach out to the
+NVIDIA Omniverse Community team at [email protected] to discuss potential opportunities
+for broader dissemination of your work.
+
+Join us in building a vibrant, collaborative ecosystem where creativity and technology intersect. Your
+contributions can make a significant impact on the Isaac Lab community and beyond!
+
 ## License
 
 The Isaac Lab framework is released under [BSD-3 License](LICENSE). The `isaaclab_mimic` extension and its corresponding standalone scripts are released under [Apache 2.0](LICENSE-mimic). The license files of its dependencies and assets are present in the [`docs/licenses`](docs/licenses) directory.

SECURITY.md

@@ -0,0 +1,38 @@
+# Security
+
+NVIDIA is dedicated to the security and trust of our software products and services, including all source code
+repositories managed through our organization.
+
+If you need to report a security issue, please use the appropriate contact points outlined below. **Please do
+not report security vulnerabilities through GitHub.**
+
+## Reporting Potential Security Vulnerability in an NVIDIA Product
+
+To report a potential security vulnerability in any NVIDIA product:
+
+- Web: [Security Vulnerability Submission Form](https://www.nvidia.com/object/submit-security-vulnerability.html)
+
+- E-Mail: [email protected]
+
+    - We encourage you to use the following PGP key for secure email communication: [NVIDIA public PGP Key for communication](https://www.nvidia.com/en-us/security/pgp-key)
+
+    - Please include the following information:
+
+    - Product/Driver name and version/branch that contains the vulnerability
+
+    - Type of vulnerability (code execution, denial of service, buffer overflow, etc.)
+
+    - Instructions to reproduce the vulnerability
+
+    - Proof-of-concept or exploit code
+
+    - Potential impact of the vulnerability, including how an attacker could exploit the vulnerability
+
+While NVIDIA currently does not have a bug bounty program, we do offer acknowledgement when an
+externally reported security issue is addressed under our coordinated vulnerability disclosure policy. Please
+visit our [Product Security Incident Response Team (PSIRT)](https://www.nvidia.com/en-us/security/psirt-policies/)
+policies page for more information.
+
+## NVIDIA Product Security
+
+For all security-related concerns, please visit NVIDIA's Product Security portal at: https://www.nvidia.com/en-us/security

source/isaaclab/config/extension.toml

@@ -1,7 +1,7 @@
 [package]
 
 # Note: Semantic Versioning is used: https://semver.org/
-version = "0.33.16"
+version = "0.34.1"
 
 # Description
 title = "Isaac Lab framework for Robot Learning"

source/isaaclab/docs/CHANGELOG.rst

@@ -1,6 +1,36 @@
 Changelog
 ---------
 
+0.34.1 (2025-02-17)
+~~~~~~~~~~~~~~~~~~~
+
+Fixed
+^^^^^
+
+* Ensured that the loaded torch JIT models inside actuator networks are correctly set to eval mode
+  to prevent any unexpected behavior during inference.
+
+
+0.34.0 (2025-02-14)
+~~~~~~~~~~~~~~~~~~~
+
+Fixed
+^^^^^
+
+* Adds attributes velocity_limits_sim and effort_limits_sim to :class:`isaaclab.actuators.AssetBaseCfg` to separate
+  solver limits from actuator limits.
+
+
+0.33.17 (2025-02-13)
+~~~~~~~~~~~~~~~~~~~~
+
+Fixed
+^^^^^
+
+* Fixed Imu sensor based observations at first step by updating scene during initialization for
+  :class:`~isaaclab.envs.ManagerBasedEnv`, :class:`~isaaclab.envs.DirectRLEnv`, and :class:`~isaaclab.envs.DirectMARLEnv`
+
+
 0.33.16 (2025-02-09)
 ~~~~~~~~~~~~~~~~~~~~
 
@@ -27,6 +57,7 @@ Fixed
 
 * Fixed not updating the timestamp of ``body_link_state_w`` and ``body_com_state_w`` when ``write_root_pose_to_sim`` and ``write_joint_state_to_sim`` in the ``Articulation`` class are called.
 
+
 0.33.13 (2025-01-30)
 ~~~~~~~~~~~~~~~~~~~~
 

source/isaaclab/isaaclab/actuators/actuator_base.py

@@ -110,6 +110,8 @@ def __init__(
         # note: for velocity limits, we don't have USD parameter, so default is infinity
         self.effort_limit = self._parse_joint_parameter(self.cfg.effort_limit, effort_limit)
         self.velocity_limit = self._parse_joint_parameter(self.cfg.velocity_limit, velocity_limit)
+        self.effort_limit_sim = self._parse_joint_parameter(self.cfg.effort_limit_sim, effort_limit)
+        self.velocity_limit_sim = self._parse_joint_parameter(self.cfg.velocity_limit_sim, velocity_limit)
 
         # create commands buffers for allocation
         self.computed_effort = torch.zeros(self._num_envs, self.num_joints, device=self._device)

source/isaaclab/isaaclab/actuators/actuator_cfg.py

@@ -33,13 +33,45 @@ class ActuatorBaseCfg:
     effort_limit: dict[str, float] | float | None = None
     """Force/Torque limit of the joints in the group. Defaults to None.
 
-    If None, the limit is set to the value specified in the USD joint prim.
+    This limit is used to clip the computed torque sent to the simulation. If None, the limit is set to the value
+    specified in the USD joint prim.
+
+    .. note::
+
+        For ImplicitActuators this value will be collapsed with effort_limit_sim due to duplicating functionality. If
+        both are set the effort_limit_sim will be used priority.
     """
 
     velocity_limit: dict[str, float] | float | None = None
     """Velocity limit of the joints in the group. Defaults to None.
 
-    If None, the limit is set to the value specified in the USD joint prim.
+    This limit is used by the actuator model. If None, the limit is set to the value specified in the USD joint prim.
+
+    .. note::
+
+        velocity_limit is not used in ActuatorBaseCfg but is provided for inherited version like
+        :class:`isaaclab.actuators.DCMotor`.
+
+    .. note::
+
+        For ImplicitActuators this value will be collapsed with velocity_limit_sim due to duplicating functionality. If
+        both are set the effort_limit_sim will be used priority.
+    """
+
+    effort_limit_sim: dict[str, float] | float | None = None
+    """Force/Torque limit of the joints in the group that will be propagated to the simulation physics solver. Defaults to None.
+
+    If None, the limit is set to the value specified in the USD joint prim for ImplicitActuators or 1.0e9 for explicit
+    actuators (e.g. IdealPDActuator). The simulation effort limits prevent computed torques from exceeding the specified
+    limit. If effort limits are too tight issues with solver convergence may occur. It is suggested to keep these value large.
+    """
+
+    velocity_limit_sim: dict[str, float] | float | None = None
+    """Velocity limit of the joints in the group that will be propagated to the simulation physics solver. Defaults to None.
+
+    If None, the limit is set to the value specified in the USD joint prim. Resulting solver solutions will constrain
+    velocities by these limits. If velocity_limit_sim is too tight issues with solver convergence may occur. It is
+    suggested to keep these value large.
     """
 
     stiffness: dict[str, float] | float | None = MISSING

source/isaaclab/isaaclab/actuators/actuator_net.py

@@ -47,7 +47,7 @@ def __init__(self, cfg: ActuatorNetLSTMCfg, *args, **kwargs):
 
         # load the model from JIT file
         file_bytes = read_file(self.cfg.network_file)
-        self.network = torch.jit.load(file_bytes, map_location=self._device)
+        self.network = torch.jit.load(file_bytes, map_location=self._device).eval()
 
         # extract number of lstm layers and hidden dim from the shape of weights
         num_layers = len(self.network.lstm.state_dict()) // 4
@@ -126,7 +126,7 @@ def __init__(self, cfg: ActuatorNetMLPCfg, *args, **kwargs):
 
         # load the model from JIT file
         file_bytes = read_file(self.cfg.network_file)
-        self.network = torch.jit.load(file_bytes, map_location=self._device)
+        self.network = torch.jit.load(file_bytes, map_location=self._device).eval()
 
         # create buffers for MLP history
         history_length = max(self.cfg.input_idx) + 1
@@ -175,7 +175,8 @@ def compute(
             )
 
         # run network inference
-        torques = self.network(network_input).view(self._num_envs, self.num_joints)
+        with torch.inference_mode():
+            torques = self.network(network_input).view(self._num_envs, self.num_joints)
         self.computed_effort = torques.view(self._num_envs, self.num_joints) * self.cfg.torque_scale
 
         # clip the computed effort based on the motor limits

source/isaaclab/isaaclab/assets/articulation/articulation.py

@@ -1339,20 +1339,48 @@ def _process_actuators_cfg(self):
             # set the passed gains and limits into the simulation
             if isinstance(actuator, ImplicitActuator):
                 self._has_implicit_actuators = True
+                # resolve actuator limit duplication for ImplicitActuators
+                # effort limits
+                if actuator.effort_limit_sim is None and actuator.effort_limit is not None:
+                    omni.log.warn(
+                        f"ImplicitActuatorCfg {actuator_name} has effort_limit_sim=None but is specifying effort_limit."
+                        "effort_limit will be applied to effort_limit_sim for ImplicitActuators."
+                    )
+                    actuator.effort_limit_sim = actuator.effort_limit
+                elif actuator.effort_limit_sim is not None and actuator.effort_limit is not None:
+                    omni.log.warn(
+                        f"ImplicitActuatorCfg {actuator_name} has set both effort_limit_sim and effort_limit."
+                        "Only effort_limit_sim will be used for ImplicitActuators."
+                    )
+                # velocity limits
+                if actuator.velocity_limit_sim is None and actuator.velocity_limit is not None:
+                    omni.log.warn(
+                        f"ImplicitActuatorCfg {actuator_name} has velocity_limit_sim=None but is specifying"
+                        " velocity_limit.velocity_limit will be applied to velocity_limit_sim for ImplicitActuators."
+                    )
+                    actuator.velocity_limit_sim = actuator.velocity_limit
+                elif actuator.velocity_limit_sim is not None and actuator.velocity_limit is not None:
+                    omni.log.warn(
+                        f"ImplicitActuatorCfg {actuator_name} has set both velocity_limit_sim and velocity_limit."
+                        "Only velocity_limit_sim will be used for ImplicitActuators."
+                    )
                 # the gains and limits are set into the simulation since actuator model is implicit
                 self.write_joint_stiffness_to_sim(actuator.stiffness, joint_ids=actuator.joint_indices)
                 self.write_joint_damping_to_sim(actuator.damping, joint_ids=actuator.joint_indices)
-                self.write_joint_effort_limit_to_sim(actuator.effort_limit, joint_ids=actuator.joint_indices)
-                self.write_joint_velocity_limit_to_sim(actuator.velocity_limit, joint_ids=actuator.joint_indices)
+                self.write_joint_effort_limit_to_sim(actuator.effort_limit_sim, joint_ids=actuator.joint_indices)
+                self.write_joint_velocity_limit_to_sim(actuator.velocity_limit_sim, joint_ids=actuator.joint_indices)
                 self.write_joint_armature_to_sim(actuator.armature, joint_ids=actuator.joint_indices)
                 self.write_joint_friction_to_sim(actuator.friction, joint_ids=actuator.joint_indices)
             else:
                 # the gains and limits are processed by the actuator model
                 # we set gains to zero, and torque limit to a high value in simulation to avoid any interference
                 self.write_joint_stiffness_to_sim(0.0, joint_ids=actuator.joint_indices)
                 self.write_joint_damping_to_sim(0.0, joint_ids=actuator.joint_indices)
-                self.write_joint_effort_limit_to_sim(1.0e9, joint_ids=actuator.joint_indices)
-                self.write_joint_velocity_limit_to_sim(actuator.velocity_limit, joint_ids=actuator.joint_indices)
+                self.write_joint_effort_limit_to_sim(
+                    1.0e9 if actuator.effort_limit_sim is None else actuator.effort_limit_sim,
+                    joint_ids=actuator.joint_indices,
+                )
+                self.write_joint_velocity_limit_to_sim(actuator.velocity_limit_sim, joint_ids=actuator.joint_indices)
                 self.write_joint_armature_to_sim(actuator.armature, joint_ids=actuator.joint_indices)
                 self.write_joint_friction_to_sim(actuator.friction, joint_ids=actuator.joint_indices)
             # Store the actual default stiffness and damping values for explicit and implicit actuators (not written the sim)

source/isaaclab/isaaclab/envs/direct_marl_env.py

@@ -133,6 +133,10 @@ def __init__(self, cfg: DirectMARLEnvCfg, render_mode: str | None = None, **kwar
             print("[INFO]: Starting the simulation. This may take a few seconds. Please wait...")
             with Timer("[INFO]: Time taken for simulation start", "simulation_start"):
                 self.sim.reset()
+                # update scene to pre populate data buffers for assets and sensors.
+                # this is needed for the observation manager to get valid tensors for initialization.
+                # this shouldn't cause an issue since later on, users do a reset over all the environments so the lazy buffers would be reset.
+                self.scene.update(dt=self.physics_dt)
 
         # -- event manager used for randomization
         if self.cfg.events:

source/isaaclab/isaaclab/envs/direct_rl_env.py

@@ -139,6 +139,10 @@ def __init__(self, cfg: DirectRLEnvCfg, render_mode: str | None = None, **kwargs
             print("[INFO]: Starting the simulation. This may take a few seconds. Please wait...")
             with Timer("[INFO]: Time taken for simulation start", "simulation_start"):
                 self.sim.reset()
+                # update scene to pre populate data buffers for assets and sensors.
+                # this is needed for the observation manager to get valid tensors for initialization.
+                # this shouldn't cause an issue since later on, users do a reset over all the environments so the lazy buffers would be reset.
+                self.scene.update(dt=self.physics_dt)
 
         # -- event manager used for randomization
         if self.cfg.events:

source/isaaclab/isaaclab/envs/manager_based_env.py

@@ -139,6 +139,10 @@ def __init__(self, cfg: ManagerBasedEnvCfg):
             print("[INFO]: Starting the simulation. This may take a few seconds. Please wait...")
             with Timer("[INFO]: Time taken for simulation start", "simulation_start"):
                 self.sim.reset()
+                # update scene to pre populate data buffers for assets and sensors.
+                # this is needed for the observation manager to get valid tensors for initialization.
+                # this shouldn't cause an issue since later on, users do a reset over all the environments so the lazy buffers would be reset.
+                self.scene.update(dt=self.physics_dt)
             # add timeline event to load managers
             self.load_managers()
 

source/isaaclab/test/assets/test_articulation.py

@@ -43,8 +43,8 @@ def generate_articulation_cfg(
     articulation_type: Literal["humanoid", "panda", "anymal", "shadow_hand", "single_joint"],
     stiffness: float | None = 10.0,
     damping: float | None = 2.0,
-    vel_limit: float | None = 100.0,
-    effort_limit: float | None = 400.0,
+    vel_limit_sim: float | None = None,
+    effort_limit_sim: float | None = None,
 ) -> ArticulationCfg:
     """Generate an articulation configuration.
 
@@ -75,8 +75,8 @@ def generate_articulation_cfg(
             actuators={
                 "joint": ImplicitActuatorCfg(
                     joint_names_expr=[".*"],
-                    effort_limit=effort_limit,
-                    velocity_limit=vel_limit,
+                    effort_limit_sim=effort_limit_sim,
+                    velocity_limit_sim=vel_limit_sim,
                     stiffness=0.0,
                     damping=10.0,
                 ),
@@ -900,7 +900,7 @@ def test_setting_gains_from_cfg_dict(self):
                         torch.testing.assert_close(articulation.actuators["body"].stiffness, expected_stiffness)
                         torch.testing.assert_close(articulation.actuators["body"].damping, expected_damping)
 
-    def test_setting_velocity_limits(self):
+    def test_setting_velocity_sim_limits(self):
         """Test that velocity limits are loaded form the configuration correctly."""
         for num_articulations in (1, 2):
             for device in ("cuda:0", "cpu"):
@@ -911,7 +911,7 @@ def test_setting_velocity_limits(self):
                         ) as sim:
                             sim._app_control_on_stop_handle = None
                             articulation_cfg = generate_articulation_cfg(
-                                articulation_type="single_joint", vel_limit=limit, effort_limit=limit
+                                articulation_type="single_joint", vel_limit_sim=limit, effort_limit_sim=limit
                             )
                             articulation, _ = generate_articulation(
                                 articulation_cfg=articulation_cfg, num_articulations=num_articulations, device=device
@@ -928,7 +928,7 @@ def test_setting_velocity_limits(self):
                                 )
                                 # Check that gains are loaded from USD file
                                 torch.testing.assert_close(
-                                    articulation.actuators["joint"].velocity_limit, expected_velocity_limit
+                                    articulation.actuators["joint"].velocity_limit_sim, expected_velocity_limit
                                 )
                                 torch.testing.assert_close(
                                     articulation.data.joint_velocity_limits, expected_velocity_limit

source/isaaclab_tasks/config/extension.toml

@@ -1,7 +1,7 @@
 [package]
 
 # Note: Semantic Versioning is used: https://semver.org/
-version = "0.10.23"
+version = "0.10.24"
 
 # Description
 title = "Isaac Lab Environments"

source/isaaclab_tasks/docs/CHANGELOG.rst

@@ -1,6 +1,15 @@
 Changelog
 ---------
 
+0.10.24 (2025-02-13)
+~~~~~~~~~~~~~~~~~~~~
+
+Changed
+^^^^^^^
+
+* Set ``Isaac-Stack-Cube-Franka-IK-Rel-v0`` to use sim parameters from base ``StackEnvCfg``, improving simulation stability.
+
+
 0.10.23 (2025-02-11)
 ~~~~~~~~~~~~~~~~~~~~
 

source/isaaclab_tasks/isaaclab_tasks/manager_based/manipulation/stack/config/franka/stack_ik_rel_env_cfg.py

@@ -34,10 +34,3 @@ def __post_init__(self):
             scale=0.5,
             body_offset=DifferentialInverseKinematicsActionCfg.OffsetCfg(pos=[0.0, 0.0, 0.107]),
         )
-
-        # Set the simulation parameters
-        self.sim.dt = 1 / 60
-        self.sim.render_interval = 1
-
-        self.decimation = 3
-        self.episode_length_s = 20.0