Velocity Limits and Protective Stop for MoveIt2 Operation

[jacob-kruse]

Hello,
I am using the MoveIt2 Python API along with the drivers in this package to motion plan with the Kuka LBR iisy 3 r760. More specifically, I am using the moveit_planning_example.launch.py and using a custom planning.yaml file to pass parameters to my MoveItPy script.

Here is a snippet from my code where I am defining the MoveIt configuration. As you can see, for ".moveit_cpp" I have defined the path to the planning.yaml:

# Define MoveIt Configuration
moveit_config = (
    MoveItConfigsBuilder('kuka_lbr_iisy')
    .robot_description(
        file_path=f'/urdf/{urdf_file}')
    .robot_description_semantic(
        file_path=(get_package_share_directory('kuka_irl_project')) +'/urdf/lbr_iisy3_r760.srdf')
    .joint_limits(
        file_path=get_package_share_directory("kuka_lbr_iisy_support")+ "/config/lbr_iisy3_r760_joint_limits.yaml")
    .moveit_cpp(
        file_path=get_package_share_directory('kuka_irl_project') + '/config/planning.yaml')
    .to_moveit_configs()
)

# Convert MoveIt Configuration to a dictionary
moveit = moveit_config.to_dict()
        lbr_iisy3_r760_joint_limits.yaml
# Instantiate MoveItPy with the manipulator component
kuka_move = MoveItPy(node_name="moveit_py", config_dict=moveit)
kuka = kuka_move.get_planning_component("manipulator")  
logger.info("MoveItPy instance created")

Here is my planning.yaml:

planning_scene_monitor_options:
  name: "planning_scene_monitor"
  robot_description: "robot_description"
  joint_state_topic: "/joint_states"
  attached_collision_object_topic: "/moveit_cpp/planning_scene_monitor"
  publish_planning_scene_topic: "/moveit_cpp/publish_planning_scene"
  monitored_planning_scene_topic: "/moveit_cpp/monitored_planning_scene"
  wait_for_initial_state_timeout: 10.0

planning_pipelines:
  pipeline_names: ["ompl", "pilz_industrial_motion_planner", 'chomp']

ompl_rrtc_solo:  # Namespace for individual plan request
  plan_request_params:moveit_cpp
    planning_attempts: 10
    planning_pipeline: ompl
    planner_id: "RRTConnectkConfigDefault"
    max_velocity_scaling_factor: 0.4
    max_acceleration_scaling_factor: 0.2
    planning_time: 2.0
  # PlanRequestParameters similar to the ones that are used by the single pipeline planning of 
pilz_lin_solo:
  plan_request_params:
    planning_attempts: 10
    planning_pipeline: pilz_industrial_motion_planner
    planner_id: "PTP"
    max_velocity_scaling_factor: 0.5
    max_acceleration_scaling_factor: 0.5
    planning_time: 2.0

chomp_planner_solo:
  plan_request_params:
    planning_attempts: 10
    planning_pipeline: chomp
    max_velocity_scaling_factor: 0.4            
    max_acceleration_scaling_factor: 0.4
    planning_time: 2.0

ompl_rrtc_multi:
  plan_request_params:
    planning_attempts: 10
    planning_pipeline: ompl
    planner_id: "RRTConnectkConfigDefault"
    max_velocity_scaling_factor: 0.4
    max_acceleration_scaling_factor: 0.4
    planning_time: 2.0

pilz_lin_multi:
  plan_request_params:
    planning_attempts: 10
    planning_pipeline: pilz_industrial_motion_planner
    planner_id: "PTP"
    max_velocity_scaling_factor: 0.4
    max_acceleration_scaling_factor: 0.4
    planning_time: 2.0

chomp_planner_multi:
  plan_request_params:
    planning_attempts: 10
    planning_pipeline: chomp
    max_velocity_scaling_factor: 0.4                
    max_acceleration_scaling_factor: 0.4
    planning_time: 2.0

With this setup, when I increase the "max_velocity_scaling_factor" and "max_acceleration_scaling_factor" too high, the robot will move at speeds that will cause the protective stop to trigger. Here is the error output when that happens:

[control_node-1] [ERROR] [1738003736.102888785] [KukaEACHardwareInterface]: External control stopped by an error
[control_node-1] [ERROR] [1738003736.102986753] [KukaEACHardwareInterface]: Session suspended
[robot_manager_node-2] [INFO] [1738003736.112866740] [robot_manager]: External control stopped
[control_node-1] [INFO] [1738003736.113318168] [resource_manager]: 'deactivate' hardware 'lbr_iisy3_r760'
[control_node-1] [INFO] [1738003736.122801288] [KukaEACHardwareInterface]: Deactivating hardware interface
[control_node-1] [INFO] [1738003736.122863807] [resource_manager]: Successful 'deactivate' of hardware 'lbr_iisy3_r760'
[control_node-1] [INFO] [1738003736.124291328] [controller_manager]: Deactivating controllers: [ joint_trajectory_controller joint_state_broadcaster ]
[robot_manager_node-2] [INFO] [1738003736.143585351] [robot_manager]: Successfully stopped controllers

I was wondering what causes this safety stop. Does the safety stop trigger when the joint velocity and joint acceleration limits defined in /kuka_lbr_iisy-support/config/lbr_iisy3_r760_joint_limits.yaml are exceeded? Assuming this is the case, if I want to increase the speed of the robot, would I have to increase the velocity and acceleration limits beyond the rated values?

[Svastits]

Could you take a look at the logbook on the controller? It is possible, that a safety velocity limit is set for the robot, which is triggered in these cases, or at high speeds/accelerations if the torque sensors are miscalibrated, a (false positive) collision can be detected.
I am pretty sure, that the error comes from safety, as in case of violating the velocity limit, some more information should be seen in the logs

[jacob-kruse]

Thanks for the quick response! This did push me in the right direction, although I should have phrased my question better. The main premise of my question was whether or not the safety configuration on the controller determined the velocity limits when I was controlling the Kuka externally (ROS2). I had disabled some safety configurations and tested, but was still finding that the Kuka was falling into protective stop. Turns out I had not applied my modified safety configuration. With the new configuration, I am able to move the Kuka at increased speeds without falling into protective stop.

Thanks again for the help,
Jacob Kruse