Request for Assistance with Robotiq Gripper Issue in Custom UR5 Environment

Isaac Sim Version : 4.1.0

Isaac Lab Version (if applicable) : 1.1.0

Operating System : Ubuntu 22.04 LTS

GPU Information

  • Model: RTX 4090
  • Driver Version: 560.35.03

Topic Description

Detailed Description

Dear NVIDIA Support Team,

I am a researcher in the field of robot learning and would like to seek your advice on an issue I am currently facing. I have created a .usd file featuring a UR5 robotic arm with a Robotiq 2F-140 gripper attached, and I am using this model to train an agent through reinforcement learning. I have also implemented a custom environment using DirectRLEnv.

However, I have encountered a problem where the gripper on the robotic arm does not behave as expected and appears to be malfunctioning. This issue seems to interfere with the training process and is also affecting the convergence of the IK solver.

I would greatly appreciate any guidance or suggestions you could provide to help resolve this issue.

Thank you for your time and assistance.

Screenshots or Videos

ArticulationCfg

The following code snippet is ArticulationCfg that I used. I hope this help you to find the solution of my issue.

UR5_CFG = ArticulationCfg(
    spawn=sim_utils.UsdFileCfg(
        usd_path="usd/ur5_with_gripper.usd",
        # usd_path = "/home/jm/workspace/ur5.usd",
        # usd_path = "usd/ur5_with_gripper_joint_disabled.usd",
        rigid_props=sim_utils.RigidBodyPropertiesCfg(
            disable_gravity=False,
            max_depenetration_velocity=5.0,
        ),
        collision_props=sim_utils.CollisionPropertiesCfg(
            collision_enabled = True,
        ),
        activate_contact_sensors=True,
    ),
    init_state=ArticulationCfg.InitialStateCfg(
        joint_pos={
            "shoulder_pan_joint": 0.0,
            "shoulder_lift_joint": -1.65,
            "elbow_joint": 1.5,
            "wrist_1_joint": -1.712,
            "wrist_2_joint": -1.712,
            "wrist_3_joint": 0.0,
            "left_inner_knuckle_joint" : 0.00033,
            "right_inner_knuckle_joint" : -0.00031,
            "left_inner_finger_joint" : 0.00051,
            "right_inner_finger_joint" : -0.00015,
            "right_outer_knuckle_joint" : -0.00034,
            "finger_joint" : 0.0
        },
        pos = (-0.2,0.0,1.0)
    ),
    actuators={
        "arm": ImplicitActuatorCfg(
            joint_names_expr=["shoulder_pan_joint", "shoulder_lift_joint", "elbow_joint", "wrist_.*_joint"],
            velocity_limit=10.0,
            effort_limit=87.0,
            stiffness=800.0,
            damping=40.0,
        ),
        "finger": ImplicitActuatorCfg(
            joint_names_expr=["finger_joint", "right_outer_knuckle_joint"],
            velocity_limit=0.2,
            effort_limit=200.0,
            stiffness=2000.0,
            damping=2000.0,
        ),
    },
)

Thank you for your interest in Isaac Lab. To ensure efficient support and collaboration, please submit your topic to its GitHub repo following the instructions provided on Isaac Lab’s Contributing Guidelines regarding discussions, submitting issues, feature requests, and contributing to the project.

We appreciate your understanding and look forward to assisting you.