Motion Control

The ROS2MotionControl module provides a high-level interface for robot motion planning and execution using MoveIt 2. It allows users to:

  • Perform Inverse Kinematics (IK) interactively by dragging a 3D gizmo in the Slicer view.

  • Plan trajectories using internal generators or MoveIt’s planning services.

  • Execute trajectories on real or simulated hardware.

Motion control builds on the robot loading workflow described in Robot Visualization. Slicer models can also be published to the MoveIt planning scene as collision objects; see the MoveIt obstacle example in Examples.

Getting Started

To use motion control, you must have a robot running with a MoveIt move_group node.

  • Launch ROS 2: Use one of the provided launch files or your own MoveIt setup. For example, for a Universal Robot:

    ros2 launch slicer_ros2_module ur_sim_control.launch.py launch_rviz:=true

  • Start Slicer: Load the robot in the ROS2 module as described in Adding and removing robots.

  • Open Motion Control: Switch to the ROS2 Motion Control module in Slicer.

Planning and execution

Interactive IK

When a robot is loaded, the Motion Control module allows you to manipulate it via a 3D transformation gizmo (sphere). Dragging this gizmo will trigger an IK request, and the robot model in Slicer will update in real-time to match the solution.

If the solver fails (e.g., the target is out of reach), the robot visualization will turn red.

Trajectory Planning

Once a goal pose is reached via IK, you can use the “Plan” button to generate a trajectory. Slicer will display a scrubber allowing you to preview the plan. Enable “Show Trajectory Path” to display the planned end-effector path as a vtkPolyLine model in the scene, and use the adjacent color selector to choose the path color.

  • Generators: Supports multiple planning backends, including MoveIt and basic joint interpolation.

  • Planning Group: Specify the MoveIt planning group (e.g., ur_manipulator) to use for planning.

Python scripts can also bypass the GUI and plan a Cartesian path through a list of Slicer 3D pose markers. Use ROS2MotionControlLogic.CreatePoseMarker to create transform-node markers, then call PlanMoveItCartesianTrajectoryFromPoseMarkers. See MoveIt Cartesian Path from Pose Markers.

After loading a planned trajectory into the module widget, scripts can call motionWidget.addPlannedTrajectoryPolyline() and motionWidget.removePlannedTrajectoryPolyline() to show or remove the same end-effector path visualization.

Execution

The “Execute” button sends the planned trajectory to the robot’s controller via the ExecuteTrajectory action.

MoveIt Scene Synchronization

The Obstacles tab allows you to use Slicer models as collision obstacles in MoveIt.

Adding Collision Objects

  • Select a Model: Choose a vtkMRMLModelNode from the dropdown.

  • Set Frame ID / TF2 Parent: Specify the ROS reference frame, usually world or base_link.

  • Track Model Transform with TF2: Keep this checked for the common workflow where a model is placed under a Slicer transform and moved with a transform gizmo or registration result.

  • Add to MoveIt: Click the Add to MoveIt button.

Once added, the obstacle is managed in the table below.

Live Synchronization

When Track Model Transform with TF2 is enabled, the UI follows the same path as ROS2MotionControlLogic.AddMoveItObstacleWithTransform. The collision object is published with header.frame_id set to the obstacle model’s Slicer node name, and the module starts a TF2 broadcaster from the selected parent frame to that obstacle frame. Moving the model’s parent transform then updates the obstacle pose in MoveIt without re-publishing the mesh.

If Track Model Transform with TF2 is disabled, the model is published as a static collision object in the frame typed in the Frame ID / TF2 Parent field. Use this mode only when the model’s polydata coordinates are already expressed in that ROS frame.

Removing Obstacles

To remove an obstacle from the MoveIt scene, simply click the Remove button in the obstacles table.