Migrate Franka Reach and Drawer to the flat asset

docs/source/_static/css/environment-browser.js

@@ -30,11 +30,11 @@
             ["Isaac-Lift-KukaAllegro-Camera", "rsl_rl", "isaacsim_physx,newton_mjwarp,ovphysx", "isaacsim_rtx,newton_renderer,ovrtx", "albedo128,albedo256,albedo64,cube,depth128,depth256,depth64,duo_camera,raycaster_depth128,raycaster_depth256,raycaster_depth64,rgb128,rgb256,rgb64,semantic_segmentation128,semantic_segmentation256,semantic_segmentation64,shapes,simple_shading_constant_diffuse128,simple_shading_constant_diffuse256,simple_shading_constant_diffuse64,simple_shading_diffuse_mdl128,simple_shading_diffuse_mdl256,simple_shading_diffuse_mdl64,simple_shading_full_mdl128,simple_shading_full_mdl256,simple_shading_full_mdl64,single_camera", "tasks/manipulation/kuka_allegro_lift.jpg", false, {"*": ["rgb64", "shapes", "single_camera"]}],
             ["Isaac-Lift-Soft-Franka", "rsl_rl", "isaacsim_physx,newton_mjwarp_vbd_proxy", "", "ik,joint", "newton/franka-mjwarp-vbd-coupling.png", false, {"*": ["joint"]}, {}, "tetrahedralization"],
             ["Isaac-Lift-Soft-Franka-Camera", "rsl_rl", "isaacsim_physx,newton_mjwarp_vbd_proxy", "isaacsim_rtx,newton_renderer,ovrtx", "ik,joint", "newton/franka-mjwarp-vbd-coupling.png", false, {"*": ["joint"]}, {}, "tetrahedralization"],
-            ["Isaac-Open-Drawer-Franka-Direct", "rl_games,rsl_rl,skrl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "", "tasks/manipulation/franka_open_drawer.jpg"],
-            ["Isaac-Open-Drawer-Franka", "rl_games,rsl_rl,skrl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "", "tasks/manipulation/franka_open_drawer.jpg"],
+            ["Isaac-Open-Drawer-Franka-Direct", "rl_games,rsl_rl,skrl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "arm_collisions", "tasks/manipulation/franka_open_drawer.jpg"],
+            ["Isaac-Open-Drawer-Franka", "rl_games,rsl_rl,skrl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "arm_collisions", "tasks/manipulation/franka_open_drawer.jpg"],
             ["Isaac-Pendulum-MARL-Direct", "rl_games,skrl", "isaacsim_physx,newton_kamino,newton_mjwarp,ovphysx", "", "", "tasks/classic/cart_double_pendulum.jpg", false, {}, {"skrl": "MAPPO"}],
-            ["Isaac-Reach-Franka", "rl_games,rsl_rl,skrl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "diffik,diffik_abs,joint_pos,newton_ik", "tasks/manipulation/franka_reach.jpg", true, {"*": ["joint_pos"]}],
-            ["Isaac-Reach-Franka-OSC", "rsl_rl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "diffik_abs", "tasks/manipulation/franka_reach.jpg", false, {"*": ["diffik_abs"]}],
+            ["Isaac-Reach-Franka", "rl_games,rsl_rl,skrl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "arm_collisions,diffik,diffik_abs,joint_pos,newton_ik", "tasks/manipulation/franka_reach.jpg", true, {"*": ["joint_pos"]}],
+            ["Isaac-Reach-Franka-OSC", "rsl_rl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "arm_collisions,diffik_abs", "tasks/manipulation/franka_reach.jpg", false, {"*": ["diffik_abs"]}],
             ["Isaac-Reach-UR10", "rl_games,rsl_rl,skrl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "", "tasks/manipulation/ur10_reach.jpg", true],
             ["Isaac-RenderBenchmark-Franka-Cabinet", "", "isaacsim_physx,newton_mjwarp,ovphysx", "isaacsim_rtx,newton_renderer,ovrtx", "albedo,depth,rgb,simple_shading_constant_diffuse,simple_shading_diffuse_mdl,simple_shading_full_mdl", "", false, {"*": ["rgb"]}],
             ["Isaac-Reorient-Cube-Allegro-Direct", "rl_games,rsl_rl,skrl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "", "tasks/manipulation/allegro_cube.jpg", true],
@@ -91,12 +91,12 @@
             ["IsaacContrib-Lift-Cube-Franka-IK-Abs", "", "", "", "", "tasks/manipulation/franka_lift.jpg"],
             ["IsaacContrib-Lift-Cube-Franka-IK-Rel", "", "", "", "", "tasks/manipulation/franka_lift.jpg"],
             ["IsaacContrib-Lift-Cube-OpenArm", "rl_games,rsl_rl", "", "", "", "tasks/manipulation/openarm_uni_lift.jpg"],
-            ["IsaacContrib-Multitask-Manipulation", "rsl_rl", "", "", "", "tasks/manipulation/multitask_manipulation.jpg"],
+            ["IsaacContrib-Multitask-Manipulation", "rsl_rl", "", "", "arm_collisions", "tasks/manipulation/multitask_manipulation.jpg"],
             ["IsaacContrib-Navigation-3DObstacles-ARL-Robot-1", "rl_games,rsl_rl,skrl", "", "", "", "tasks/drone_arl/arl_robot_1_navigation.jpg"],
             ["IsaacContrib-Navigation-Flat-AnymalC", "rsl_rl,skrl", "", "", "", "tasks/navigation/anymal_c_nav.jpg"],
             ["IsaacContrib-NutPour-GR1T2-Pink-IK-Abs", "", "", "", ""],
-            ["IsaacContrib-Open-Drawer-Franka-IK-Abs", "rsl_rl", "isaacsim_physx,newton_mjwarp,ovphysx", "", ""],
-            ["IsaacContrib-Open-Drawer-Franka-IK-Rel", "rsl_rl", "isaacsim_physx,newton_mjwarp,ovphysx", "", ""],
+            ["IsaacContrib-Open-Drawer-Franka-IK-Abs", "rsl_rl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "arm_collisions"],
+            ["IsaacContrib-Open-Drawer-Franka-IK-Rel", "rsl_rl", "isaacsim_physx,newton_mjwarp,ovphysx", "", "arm_collisions"],
             ["IsaacContrib-Open-Drawer-OpenArm", "rl_games,rsl_rl", "", "", "", "tasks/manipulation/openarm_uni_open_drawer.jpg"],
             ["IsaacContrib-PickPlace-FixedBaseUpperBodyIK-G1-Abs", "", "", "isaacsim_rtx,newton_renderer,ovrtx", "", "tasks/manipulation/g1_pick_place_fixed_base.jpg"],
             ["IsaacContrib-PickPlace-G1-InspireFTP-Abs", "", "", "", "", "tasks/manipulation/g1_pick_place.jpg"],

examples/haply_teleoperation.py

@@ -33,7 +33,7 @@
 from isaaclab.utils import configclass, instantiate, replace
 from isaaclab.utils.assets import ISAAC_NUCLEUS_DIR
 
-from isaaclab_assets import FRANKA_PANDA_HIGH_PD_CFG
+from isaaclab_assets import FRANKA_PANDA_FLAT_HIGH_PD_CFG
 
 if TYPE_CHECKING:
     from isaaclab.assets import Articulation, RigidObject
@@ -127,10 +127,11 @@ class FrankaHaplySceneCfg(InteractiveSceneCfg):
         init_state=AssetBaseCfg.InitialStateCfg(pos=(0.50, 0.0, 1.05), rot=(0.707, 0, 0, 0.707)),
     )
 
-    robot_cfg: ArticulationCfg = replace(FRANKA_PANDA_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
+    robot_cfg: ArticulationCfg = replace(FRANKA_PANDA_FLAT_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
     robot = instantiate(robot_cfg)
     robot.init_state.pos = (-0.02, 0.0, 1.05)
     robot.spawn.activate_contact_sensors = True
+    robot.spawn.variants["Physics"] = "mujoco" if args_cli.physics == "newton_mjwarp" else "physx"
 
     cube = RigidObjectCfg(
         prim_path="{ENV_REGEX_NS}/Cube",
@@ -146,15 +147,15 @@ class FrankaHaplySceneCfg(InteractiveSceneCfg):
     )
 
     left_finger_contact_sensor = ContactSensorCfg(
-        prim_path="{ENV_REGEX_NS}/Robot/panda_leftfinger",
+        prim_path="{ENV_REGEX_NS}/Robot/(Geometry/.*/)?panda_leftfinger",
         update_period=0.0,
         history_length=3,
         debug_vis=True,
         track_pose=True,
     )
 
     right_finger_contact_sensor = ContactSensorCfg(
-        prim_path="{ENV_REGEX_NS}/Robot/panda_rightfinger",
+        prim_path="{ENV_REGEX_NS}/Robot/(Geometry/.*/)?panda_rightfinger",
         update_period=0.0,
         history_length=3,
         debug_vis=True,

scripts/tutorials/05_controllers/run_diff_ik.py

@@ -53,7 +53,7 @@
 ##
 # Pre-defined configs
 ##
-from isaaclab_assets import FRANKA_PANDA_HIGH_PD_CFG, UR10_CFG  # isort:skip
+from isaaclab_assets import FRANKA_PANDA_FLAT_HIGH_PD_CFG, UR10_CFG  # isort:skip
 
 
 @configclass
@@ -82,7 +82,7 @@ class TableTopSceneCfg(InteractiveSceneCfg):
 
     # articulation
     if args_cli.robot == "franka_panda":
-        robot = replace(FRANKA_PANDA_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
+        robot = replace(FRANKA_PANDA_FLAT_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
     elif args_cli.robot == "ur10":
         robot = replace(UR10_CFG, prim_path="{ENV_REGEX_NS}/Robot")
     else:

scripts/tutorials/05_controllers/run_osc.py

@@ -57,7 +57,7 @@
 ##
 # Pre-defined configs
 ##
-from isaaclab_assets import FRANKA_PANDA_HIGH_PD_CFG  # isort:skip
+from isaaclab_assets import FRANKA_PANDA_FLAT_HIGH_PD_CFG  # isort:skip
 
 
 @configclass
@@ -97,11 +97,9 @@ class SceneCfg(InteractiveSceneCfg):
         debug_vis=False,
     )
 
-    robot = replace(FRANKA_PANDA_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
-    robot.actuators["panda_shoulder"].stiffness = 0.0
-    robot.actuators["panda_shoulder"].damping = 0.0
-    robot.actuators["panda_forearm"].stiffness = 0.0
-    robot.actuators["panda_forearm"].damping = 0.0
+    robot = replace(FRANKA_PANDA_FLAT_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
+    robot.actuators["panda_arm"].stiffness = 0.0
+    robot.actuators["panda_arm"].damping = 0.0
     robot.spawn.rigid_props.disable_gravity = True
 
 

source/isaaclab/changelog.d/maximiliank-diffik-action-offset.minor.rst

@@ -0,0 +1,5 @@
+Added
+^^^^^
+
+* Added an ``offset`` to ``DifferentialInverseKinematicsActionCfg`` so normalized policies can use
+  an affine task-space action transform. The default is zero and preserves existing configurations.

source/isaaclab/isaaclab/envs/mdp/actions/actions_cfg.py

@@ -325,6 +325,8 @@ class OffsetCfg:
     """Offset of target frame w.r.t. to the body frame. Defaults to None, in which case no offset is applied."""
     scale: float | tuple[float, ...] = 1.0
     """Scale factor for the action. Defaults to 1.0."""
+    offset: float | tuple[float, ...] = 0.0
+    """Offset applied to the scaled action in controller-specific units. Defaults to 0.0."""
     controller: DifferentialIKControllerCfg = MISSING
     """The configuration for the differential IK controller."""
 

source/isaaclab/isaaclab/envs/mdp/actions/task_space_actions.py

@@ -34,23 +34,25 @@
 class DifferentialInverseKinematicsAction(ActionTerm):
     r"""Inverse Kinematics action term.
 
-    This action term performs pre-processing of the raw actions using scaling transformation.
+    This action term pre-processes raw actions using an affine transformation.
 
     .. math::
-        \text{action} = \text{scaling} \times \text{input action}
+        \text{action} = \text{offset} + \text{scaling} \times \text{input action}
         \text{joint position} = J^{-} \times \text{action}
 
-    where :math:`\text{scaling}` is the scaling applied to the input action, and :math:`\text{input action}`
-    is the input action from the user, :math:`J` is the Jacobian over the articulation's actuated joints,
-    and \text{joint position} is the desired joint position command for the articulation's joints.
+    where :math:`\text{offset}` and :math:`\text{scaling}` define the affine transformation,
+    :math:`\text{input action}` is the input action from the user, :math:`J` is the Jacobian over the
+    articulation's actuated joints, and \text{joint position} is the desired joint position command.
     """
 
     cfg: actions_cfg.DifferentialInverseKinematicsActionCfg
     """The configuration of the action term."""
     _asset: Articulation
     """The articulation asset on which the action term is applied."""
     _scale: torch.Tensor
-    """The scaling factor applied to the input action. Shape is (1, action_dim)."""
+    """The scaling factor applied to the input action. Shape is (num_envs, action_dim)."""
+    _offset: torch.Tensor
+    """The offset applied to the scaled action. Shape is (num_envs, action_dim)."""
     _clip: torch.Tensor
     """The clip applied to the input action."""
 
@@ -100,9 +102,11 @@ def __init__(self, cfg: actions_cfg.DifferentialInverseKinematicsActionCfg, env:
         # owned buffer; _compute_frame_jacobian mutates this, not the data-layer view.
         self._jacobian_b = torch.zeros(self.num_envs, 6, len(self._jacobi_joint_ids), device=self.device)
 
-        # save the scale as tensors
+        # save the affine transform as tensors
         self._scale = torch.zeros((self.num_envs, self.action_dim), device=self.device)
         self._scale[:] = torch.tensor(self.cfg.scale, device=self.device)
+        self._offset = torch.zeros((self.num_envs, self.action_dim), device=self.device)
+        self._offset[:] = torch.tensor(self.cfg.offset, device=self.device)
 
         # convert the fixed offsets to torch tensors of batched shape
         if self.cfg.body_offset is not None:
@@ -160,6 +164,7 @@ def IO_descriptor(self) -> GenericActionIODescriptor:
         - body_name: The name of the body.
         - joint_names: The names of the joints.
         - scale: The scale of the action term.
+        - offset: The offset of the action term.
         - clip: The clip of the action term.
         - controller_cfg: The configuration of the controller.
         - body_offset: The offset of the body.
@@ -174,6 +179,7 @@ def IO_descriptor(self) -> GenericActionIODescriptor:
         self._IO_descriptor.body_name = self._body_name
         self._IO_descriptor.joint_names = self._joint_names
         self._IO_descriptor.scale = self._scale
+        self._IO_descriptor.offset = self._offset
         if self.cfg.clip is not None:
             self._IO_descriptor.clip = self.cfg.clip
         else:
@@ -189,7 +195,7 @@ def IO_descriptor(self) -> GenericActionIODescriptor:
     def process_actions(self, actions: torch.Tensor):
         # store the raw actions
         self._raw_actions[:] = actions
-        self._processed_actions[:] = self.raw_actions * self._scale
+        self._processed_actions[:] = self.raw_actions * self._scale + self._offset
         if self.cfg.clip is not None:
             self._processed_actions = torch.clamp(
                 self._processed_actions, min=self._clip[:, :, 0], max=self._clip[:, :, 1]

source/isaaclab/test/controllers/test_differential_ik.py

@@ -31,7 +31,7 @@
 ##
 # Pre-defined configs
 ##
-from isaaclab_assets import FRANKA_PANDA_HIGH_PD_CFG, UR10_CFG  # isort:skip
+from isaaclab_assets import FRANKA_PANDA_FLAT_HIGH_PD_CFG, UR10_CFG  # isort:skip
 
 pytestmark = pytest.mark.integration
 
@@ -441,7 +441,7 @@ def test_franka_ik_pose_abs(sim):
     sim_context, num_envs, ee_pose_b_des_set = sim
 
     # Create robot instance
-    robot_cfg = replace(FRANKA_PANDA_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
+    robot_cfg = replace(FRANKA_PANDA_FLAT_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
     robot = Articulation(cfg=robot_cfg)
 
     # Create IK controller

source/isaaclab/test/controllers/test_operational_space.py

@@ -50,7 +50,7 @@
     subtract_frame_transforms,
 )
 
-from isaaclab_assets import FRANKA_PANDA_CFG, G1_29DOF_CFG  # isort:skip
+from isaaclab_assets import FRANKA_PANDA_LEGACY_CFG, G1_29DOF_CFG  # isort:skip
 
 pytestmark = pytest.mark.integration
 
@@ -95,7 +95,8 @@ def sim():
     # clone the env xform
     cloner.usd_replicate(stage, [env_fmt.format(0)], [env_fmt], env_ids, positions=env_origins)
 
-    robot_cfg = replace(FRANKA_PANDA_CFG, prim_path="{ENV_REGEX_NS}/Robot")
+    # Keep controller regressions on their original plant; canonical asset parity is tested by each backend.
+    robot_cfg = replace(FRANKA_PANDA_LEGACY_CFG, prim_path="{ENV_REGEX_NS}/Robot")
     # Explicit torque actuators enforce effort limits on the commands sent to the simulator.
     for actuator_name in ("panda_shoulder", "panda_forearm"):
         actuator_cfg = robot_cfg.actuators[actuator_name]

source/isaaclab/test/envs/test_diffik_action_processing.py

@@ -0,0 +1,56 @@
+# Copyright (c) 2022-2026, The Isaac Lab Project Developers (https://github.com/isaac-sim/IsaacLab/blob/main/CONTRIBUTORS.md).
+# All rights reserved.
+#
+# SPDX-License-Identifier: BSD-3-Clause
+
+"""Unit tests for differential inverse-kinematics action processing."""
+
+from types import SimpleNamespace
+
+import pytest
+import torch
+
+from isaaclab.controllers import DifferentialIKControllerCfg
+from isaaclab.envs.mdp.actions.actions_cfg import DifferentialInverseKinematicsActionCfg
+from isaaclab.envs.mdp.actions.task_space_actions import DifferentialInverseKinematicsAction
+
+pytestmark = pytest.mark.unit
+
+
+def test_process_actions_applies_scale_then_offset(monkeypatch: pytest.MonkeyPatch) -> None:
+    """DiffIK actions use the configured affine transformation before reaching the controller."""
+    num_envs = 2
+    controller_cfg = DifferentialIKControllerCfg(command_type="position", use_relative_mode=False, ik_method="dls")
+    cfg = DifferentialInverseKinematicsActionCfg(
+        asset_name="robot",
+        joint_names=[".*"],
+        body_name="tool",
+        scale=(0.5, 2.0, -1.0),
+        offset=(0.1, -0.2, 0.3),
+        controller=controller_cfg,
+    )
+    asset = SimpleNamespace(
+        find_joints=lambda _names: ([0, 1, 2], ["joint1", "joint2", "joint3"]),
+        find_bodies=lambda _name: ([0], ["tool"]),
+        num_joints=3,
+        num_base_dofs=0,
+        is_fixed_base=False,
+    )
+    env = SimpleNamespace(scene={"robot": asset}, num_envs=num_envs, device="cpu")
+    action = DifferentialInverseKinematicsAction(cfg, env)
+    ee_pos = torch.zeros(num_envs, 3)
+    ee_quat = torch.tensor([[0.0, 0.0, 0.0, 1.0]]).repeat(num_envs, 1)
+    monkeypatch.setattr(action, "_compute_frame_pose", lambda: (ee_pos, ee_quat))
+
+    raw_actions = torch.tensor([[1.0, -2.0, 0.5], [-1.0, 0.25, -0.5]])
+    expected = torch.tensor([[0.6, -4.2, -0.2], [-0.4, 0.3, 0.8]])
+
+    action.process_actions(raw_actions)
+
+    torch.testing.assert_close(action.raw_actions, raw_actions)
+    torch.testing.assert_close(action.processed_actions, expected)
+    torch.testing.assert_close(action._ik_controller.ee_pos_des, expected)
+
+
+if __name__ == "__main__":
+    pytest.main([__file__, "-v"])

source/isaaclab_tasks/changelog.d/maximiliank-franka-reach-drawer.major.rst

@@ -0,0 +1,11 @@
+Changed
+^^^^^^^
+
+* Migrated maintained Franka Reach, Drawer, and related contributed tasks to the shared flat asset with
+  backend-specific physics and gripper-only collisions by default. Select ``arm_collisions`` where arm
+  contacts are required. Existing checkpoints should be requalified against the changed robot dynamics.
+
+Fixed
+^^^^^
+
+* Corrected the Reach action and controller contracts for the shared asset.

source/isaaclab_tasks/isaaclab_tasks/benchmark/render_benchmark/render_benchmark_env_cfg.py

@@ -25,10 +25,10 @@
 from isaaclab.utils import configclass, replace
 from isaaclab.utils.assets import ISAAC_NUCLEUS_DIR
 
-from isaaclab_tasks.utils import PresetCfg
+from isaaclab_tasks.utils import PresetCfg, preset
 from isaaclab_tasks.utils.presets import MultiBackendRendererCfg
 
-from isaaclab_assets.robots.franka import FRANKA_PANDA_HIGH_PD_CFG
+from isaaclab_assets.robots.franka import FRANKA_PANDA_FLAT_HIGH_PD_CFG
 
 BenchmarkMode = Literal["render", "physics_render"]
 """Animation mode used by the render benchmark.
@@ -144,10 +144,11 @@ class RenderBenchmarkSceneCfg(InteractiveSceneCfg):
         init_state=RigidObjectCfg.InitialStateCfg(pos=(0.0, 0.0, -0.05)),
     )
     robot: ArticulationCfg = replace(
-        FRANKA_PANDA_HIGH_PD_CFG,
+        FRANKA_PANDA_FLAT_HIGH_PD_CFG,
         prim_path="{ENV_REGEX_NS}/Robot",
-        init_state=replace(FRANKA_PANDA_HIGH_PD_CFG.init_state, pos=(1.0, 0.0, 0.0), rot=(0.0, 0.0, 1.0, 0.0)),
+        init_state=replace(FRANKA_PANDA_FLAT_HIGH_PD_CFG.init_state, pos=(1.0, 0.0, 0.0), rot=(0.0, 0.0, 1.0, 0.0)),
     )
+    robot.spawn.variants["Physics"] = preset(default="mujoco", isaacsim_physx="physx", ovphysx="physx", physx="physx")
     cabinet: ArticulationCfg = ArticulationCfg(
         prim_path="{ENV_REGEX_NS}/Cabinet",
         spawn=sim_utils.UsdFileCfg(

source/isaaclab_tasks/isaaclab_tasks/contrib/cabinet/config/franka/ik_abs_env_cfg.py

@@ -12,7 +12,7 @@
 ##
 # Pre-defined configs
 ##
-from isaaclab_assets.robots.franka import FRANKA_PANDA_HIGH_PD_CFG  # isort: skip
+from isaaclab_assets.robots.franka import FRANKA_PANDA_FLAT_HIGH_PD_CFG  # isort: skip
 
 
 @configclass
@@ -23,7 +23,9 @@ def __post_init__(self):
 
         # Set Franka as robot
         # We switch here to a stiffer PD controller for IK tracking to be better.
-        self.scene.robot = replace(FRANKA_PANDA_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
+        physics_variants = self.scene.robot.spawn.variants
+        self.scene.robot = replace(FRANKA_PANDA_FLAT_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
+        self.scene.robot.spawn.variants = physics_variants
 
         # Set actions for the specific robot type (franka)
         self.actions.arm_action = DifferentialInverseKinematicsActionCfg(

source/isaaclab_tasks/isaaclab_tasks/contrib/cabinet/config/franka/ik_rel_env_cfg.py

@@ -12,7 +12,7 @@
 ##
 # Pre-defined configs
 ##
-from isaaclab_assets.robots.franka import FRANKA_PANDA_HIGH_PD_CFG  # isort: skip
+from isaaclab_assets.robots.franka import FRANKA_PANDA_FLAT_HIGH_PD_CFG  # isort: skip
 
 
 @configclass
@@ -23,7 +23,9 @@ def __post_init__(self):
 
         # Set Franka as robot
         # We switch here to a stiffer PD controller for IK tracking to be better.
-        self.scene.robot = replace(FRANKA_PANDA_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
+        physics_variants = self.scene.robot.spawn.variants
+        self.scene.robot = replace(FRANKA_PANDA_FLAT_HIGH_PD_CFG, prim_path="{ENV_REGEX_NS}/Robot")
+        self.scene.robot.spawn.variants = physics_variants
 
         # Set actions for the specific robot type (franka)
         self.actions.arm_action = DifferentialInverseKinematicsActionCfg(