[New Robot] OpenLoong Support

README.md

@@ -63,6 +63,15 @@ python scripts/rsl_rl/train.py --task RobotLab-Isaac-Velocity-Flat-Unitree-H1-v0
 python scripts/rsl_rl/play.py --task RobotLab-Isaac-Velocity-Flat-Unitree-H1-v0
 ```
 
+OpenLoong OpenLoong
+
+```bash
+# Train
+python scripts/rsl_rl/train.py --task RobotLab-Isaac-Velocity-Flat-OpenLoong-OpenLoong-v0 --headless
+# Play
+python scripts/rsl_rl/play.py --task RobotLab-Isaac-Velocity-Flat-OpenLoong-OpenLoong-v0
+```
+
 The above configs are flat, you can change Flat to Rough
 
 **Note**

exts/robot_lab/robot_lab/assets/openloong.py

@@ -0,0 +1,81 @@
+"""Configuration for OPENLOONG robots.
+
+The following configurations are available:
+
+* :obj:`OPENLOONG_OPENLOONG_CFG`: OPENLOONG OPENLOONG humanoid robot
+
+Reference: https://www.openloong.org.cn/cn
+"""
+
+from robot_lab.assets import ISAACLAB_ASSETS_DATA_DIR
+
+import omni.isaac.lab.sim as sim_utils
+from omni.isaac.lab.actuators import ImplicitActuatorCfg
+from omni.isaac.lab.assets.articulation import ArticulationCfg
+
+##
+# Configuration
+##
+
+
+OPENLOONG_OPENLOONG_CFG = ArticulationCfg(
+    spawn=sim_utils.UsdFileCfg(
+        usd_path=f"{ISAACLAB_ASSETS_DATA_DIR}/Robots/OpenLoong/OpenLoong.usd",
+        activate_contact_sensors=True,
+        rigid_props=sim_utils.RigidBodyPropertiesCfg(
+            disable_gravity=False,
+            retain_accelerations=False,
+            linear_damping=0.0,
+            angular_damping=0.0,
+            max_linear_velocity=1000.0,
+            max_angular_velocity=1000.0,
+            max_depenetration_velocity=1.0,
+        ),
+        articulation_props=sim_utils.ArticulationRootPropertiesCfg(
+            enabled_self_collisions=False, solver_position_iteration_count=4, solver_velocity_iteration_count=4
+        ),
+    ),
+    init_state=ArticulationCfg.InitialStateCfg(
+        pos=(0.0, 0.0, 1.09),
+        joint_pos={
+            #left_leg
+           'l_hip_roll' : 0.,  
+           'l_hip_yaw' : 0. ,   
+           'l_hip_pitch' : 0.,         
+           'l_knee_pitch' : 0.0,       
+           'l_ankle_pitch': 0.,
+           'l_ankle_roll': 0.,  
+           #right_leg  
+           'r_hip_roll' : 0., 
+           'r_hip_yaw' : 0., 
+           'r_hip_pitch' : 0.,                                       
+           'r_knee_pitch' : 0.0,                                             
+           'r_ankle_pitch': 0.,
+           'r_ankle_roll': 0.,
+        },
+        joint_vel={".*": 0.0},
+    ),
+    soft_joint_pos_limit_factor=0.9,
+    actuators={
+        "actuators": ImplicitActuatorCfg(
+            joint_names_expr=[".*"],
+            stiffness={
+                '.*_hip_roll': 251.625,
+                '.*_hip_pitch': 200,
+                '.*_hip_yaw': 362.5214,
+                '.*_knee_pitch': 200,
+                '.*_ankle_pitch': 10.9805,
+                '.*_ankle_roll': 10.9805,
+            },
+            damping={
+                '.*_hip_roll': 10,
+                '.*_hip_pitch': 10,
+                '.*_hip_yaw': 10,
+                '.*_knee_pitch':10,
+                '.*_ankle_pitch': 10,
+                '.*_ankle_roll': 10,                   
+            },
+        ),
+    },
+)
+"""Configuration for the OPENLOONG OPENLOONG Humanoid robot."""

exts/robot_lab/robot_lab/tasks/locomotion/velocity/config/openloong/__init__.py

@@ -0,0 +1,27 @@
+import gymnasium as gym
+
+from . import agents, flat_env_cfg, rough_env_cfg
+
+##
+# Register Gym environments.
+##
+
+gym.register(
+    id="RobotLab-Isaac-Velocity-Rough-OpenLoong-OpenLoong-v0",
+    entry_point="omni.isaac.lab.envs:ManagerBasedRLEnv",
+    disable_env_checker=True,
+    kwargs={
+        "env_cfg_entry_point": rough_env_cfg.OpenLoongOpenLoongRoughEnvCfg,
+        "rsl_rl_cfg_entry_point": f"{agents.__name__}.rsl_rl_ppo_cfg:OpenLoongOpenLoongRoughPPORunnerCfg",
+    },
+)
+
+gym.register(
+    id="RobotLab-Isaac-Velocity-Flat-OpenLoong-OpenLoong-v0",
+    entry_point="omni.isaac.lab.envs:ManagerBasedRLEnv",
+    disable_env_checker=True,
+    kwargs={
+        "env_cfg_entry_point": flat_env_cfg.OpenLoongOpenLoongFlatEnvCfg,
+        "rsl_rl_cfg_entry_point": f"{agents.__name__}.rsl_rl_ppo_cfg:OpenLoongOpenLoongFlatPPORunnerCfg",
+    },
+)

exts/robot_lab/robot_lab/tasks/locomotion/velocity/config/openloong/agents/rsl_rl_ppo_cfg.py

@@ -0,0 +1,46 @@
+from omni.isaac.lab.utils import configclass
+from omni.isaac.lab_tasks.utils.wrappers.rsl_rl import (
+    RslRlOnPolicyRunnerCfg,
+    RslRlPpoActorCriticCfg,
+    RslRlPpoAlgorithmCfg,
+)
+
+
+@configclass
+class OpenLoongOpenLoongRoughPPORunnerCfg(RslRlOnPolicyRunnerCfg):
+    num_steps_per_env = 24
+    max_iterations = 3000
+    save_interval = 50
+    experiment_name = "openloong_rough"
+    empirical_normalization = False
+    policy = RslRlPpoActorCriticCfg(
+        init_noise_std=1.0,
+        actor_hidden_dims=[512, 256, 128],
+        critic_hidden_dims=[512, 256, 128],
+        activation="elu",
+    )
+    algorithm = RslRlPpoAlgorithmCfg(
+        value_loss_coef=1.0,
+        use_clipped_value_loss=True,
+        clip_param=0.2,
+        entropy_coef=0.01,
+        num_learning_epochs=5,
+        num_mini_batches=4,
+        learning_rate=1.0e-3,
+        schedule="adaptive",
+        gamma=0.99,
+        lam=0.95,
+        desired_kl=0.01,
+        max_grad_norm=1.0,
+    )
+
+
+@configclass
+class OpenLoongOpenLoongFlatPPORunnerCfg(OpenLoongOpenLoongRoughPPORunnerCfg):
+    def __post_init__(self):
+        super().__post_init__()
+
+        self.max_iterations = 1000
+        self.experiment_name = "openloong_flat"
+        self.policy.actor_hidden_dims = [128, 128, 128]
+        self.policy.critic_hidden_dims = [128, 128, 128]

exts/robot_lab/robot_lab/tasks/locomotion/velocity/config/openloong/flat_env_cfg.py

@@ -0,0 +1,29 @@
+from omni.isaac.lab.utils import configclass
+
+from .rough_env_cfg import OpenLoongOpenLoongRoughEnvCfg
+
+
+@configclass
+class OpenLoongOpenLoongFlatEnvCfg(OpenLoongOpenLoongRoughEnvCfg):
+    def __post_init__(self):
+        # Temporarily not run disable_zerow_eight_rewards() in parent class to override rewards
+        self._run_disable_zero_weight_rewards = False
+        # post init of parent
+        super().__post_init__()
+
+        # override rewards
+        self.rewards.feet_air_time.weight = 1.0
+        self.rewards.feet_air_time.params["threshold"] = 0.5
+        # change terrain to flat
+        self.scene.terrain.terrain_type = "plane"
+        self.scene.terrain.terrain_generator = None
+        # no height scan
+        self.scene.height_scanner = None
+        self.observations.policy.height_scan = None
+        # no terrain curriculum
+        self.curriculum.terrain_levels = None
+
+        # Now executing disable_zerow_eight_rewards()
+        self._run_disable_zero_weight_rewards = True
+        if self._run_disable_zero_weight_rewards:
+            self.disable_zero_weight_rewards()

exts/robot_lab/robot_lab/tasks/locomotion/velocity/config/openloong/rough_env_cfg.py

@@ -0,0 +1,128 @@
+import robot_lab.tasks.locomotion.velocity.mdp as mdp
+from robot_lab.tasks.locomotion.velocity.velocity_env_cfg import LocomotionVelocityRoughEnvCfg
+
+from omni.isaac.lab.utils import configclass
+
+##
+# Pre-defined configs
+##
+# use local assets
+from robot_lab.assets.openloong import OPENLOONG_OPENLOONG_CFG # isort: skip
+
+
+@configclass
+class OpenLoongOpenLoongRoughEnvCfg(LocomotionVelocityRoughEnvCfg):
+    _run_disable_zero_weight_rewards = True
+
+    def __post_init__(self):
+        # post init of parent
+        super().__post_init__()
+
+        # ------------------------------Sence------------------------------
+        # switch robot to fftai-GR1T1
+        self.scene.robot = OPENLOONG_OPENLOONG_CFG.replace(prim_path="{ENV_REGEX_NS}/Robot")
+        self.scene.height_scanner.prim_path = "{ENV_REGEX_NS}/Robot/base_link"
+
+        # ------------------------------Observations------------------------------
+        # self.observations.policy.base_lin_vel = None
+        # self.observations.policy.height_scan = None
+        self.observations.AMP = None
+        # ------------------------------Actions------------------------------
+        # reduce action scale
+        # self.actions.joint_pos.scale = 1.0
+
+        # ------------------------------Events------------------------------
+        self.events.reset_base_amp = None
+        self.events.reset_robot_joints_amp = None
+        self.events.base_external_force_torque.params["asset_cfg"].body_names = ["base_link"]
+        self.events.reset_robot_joints.params["position_range"] = (1.0, 1.0)
+        self.events.reset_base.params = {
+            "pose_range": {"x": (-0.5, 0.5), "y": (-0.5, 0.5), "yaw": (-3.14, 3.14)},
+            "velocity_range": {
+                "x": (0.0, 0.0),
+                "y": (0.0, 0.0),
+                "z": (0.0, 0.0),
+                "roll": (0.0, 0.0),
+                "pitch": (0.0, 0.0),
+                "yaw": (0.0, 0.0),
+            },
+        }
+        self.events.push_robot = None
+        self.events.add_base_mass = None
+        self.events.randomize_actuator_gains = None
+        self.events.randomize_joint_parameters = None
+
+        # ------------------------------Rewards------------------------------
+        # General
+        # UNUESD self.rewards.is_alive.weight = 0
+        self.rewards.is_terminated.weight = -200
+
+        # Root penalties
+        self.rewards.lin_vel_z_l2.weight = 0
+        self.rewards.ang_vel_xy_l2.weight = -0.05
+        self.rewards.flat_orientation_l2.weight = -1.0
+        self.rewards.base_height_l2.weight = 0
+        self.rewards.body_lin_acc_l2.weight = 0
+
+        # Joint penaltie
+        self.rewards.joint_torques_l2.weight = 0
+        # UNUESD self.rewards.joint_vel_l1.weight = 0.0
+        self.rewards.joint_vel_l2.weight = 0
+        self.rewards.joint_acc_l2.weight = -1.25e-7
+        self.rewards.create_joint_deviation_l1_rewterm(
+            "joint_deviation_other_l1", -1.0, [".*_hip_yaw", ".*_hip_roll"]
+        )
+        self.rewards.create_joint_deviation_l1_rewterm("joint_deviation_knee_l1", -0.1, ['.*_knee_pitch'])
+        self.rewards.joint_pos_limits.weight = -1.0
+        self.rewards.joint_pos_limits.params["asset_cfg"].joint_names = [".*_ankle_pitch",".*_ankle_roll"]
+        self.rewards.joint_vel_limits.weight = 0
+
+        # Action penalties
+        self.rewards.applied_torque_limits.weight = 0
+        self.rewards.action_rate_l2.weight = -0.005
+        # UNUESD self.rewards.action_l2.weight = 0.0
+
+        # Contact sensor
+        self.rewards.undesired_contacts.weight = 0
+        self.rewards.contact_forces.weight = 0
+
+        # Velocity-tracking rewards
+        self.rewards.track_lin_vel_xy_exp.weight = 1.0
+        self.rewards.track_lin_vel_xy_exp.func = mdp.track_lin_vel_xy_yaw_frame_exp
+        self.rewards.track_lin_vel_xy_exp.params["std"] = 0.5
+        self.rewards.track_ang_vel_z_exp.weight = 1.0
+        self.rewards.track_ang_vel_z_exp.func = mdp.track_ang_vel_z_world_exp
+        self.rewards.track_ang_vel_z_exp.params["std"] = 0.5
+
+        # Others
+        self.rewards.feet_air_time.weight = 0.25
+        self.rewards.feet_air_time.func = mdp.feet_air_time_positive_biped
+        self.rewards.feet_air_time.params["sensor_cfg"].body_names = ["Link_r_ankle_roll","Link_l_ankle_roll"]
+        self.rewards.feet_air_time.params["threshold"] = 0.4
+        self.rewards.foot_contact.weight = 0
+        self.rewards.base_height_rough_l2.weight = 0
+        self.rewards.feet_slide.weight = -0.25
+        self.rewards.feet_slide.params["sensor_cfg"].body_names = ["Link_r_ankle_roll","Link_l_ankle_roll"]
+        self.rewards.feet_slide.params["asset_cfg"].body_names = ["Link_r_ankle_roll","Link_l_ankle_roll"]
+        self.rewards.joint_power.weight = 0
+        self.rewards.stand_still_when_zero_command.weight = 0
+
+        # If the weight of rewards is 0, set rewards to None
+        if self._run_disable_zero_weight_rewards:
+            self.disable_zero_weight_rewards()
+
+        # ------------------------------Terminations------------------------------
+        self.terminations.illegal_contact.params["sensor_cfg"].body_names = [
+            "base_link",
+            ".*_waist_.*",
+            ".*_head_.*",
+            ".*_arm_r_.*",
+            ".*_arm_l_.*",
+            ".*_hip_r_.*",
+            ".*_hip_l_.*"
+        ]
+
+        # ------------------------------Commands------------------------------
+        self.commands.base_velocity.ranges.lin_vel_x = (0.0, 1.0)
+        self.commands.base_velocity.ranges.lin_vel_y = (0.0, 0.0)
+        self.commands.base_velocity.ranges.ang_vel_z = (-1.0, 1.0)