ComputePlan

This state allows you to compute a plan between two states of the robot. This is a crucial component in making a robot move from a point A to a point B.

Outcomes

This state has two outcomes: success (green socket) and failure (red socket). If the configuration of the state is wrong, or if a motion plan cannot be computed for the designated group, failure is returned. Return the other outcome if the operation is successful.

How to configure

Once dropped in the task editor, you should see five or six parameters appear in the state:

  • group_name: Name of the MoveIt! group for which the plan will be computed. This slot will not appear if only one group is defined.

  • target_type: Name of the manager from which the target state should be retrieved. Valid values are joint state and pose. If left empty, the target will be retrieved from the userdata.

  • target_name: Name of the target state. It can be the name of either a pose or a joint state which is located either in a manager or in the userdata.

  • plan_name: Name to give to the plan. If left empty stores the resulting plan in the commander as the last computed one. If not left empty, sends the computed plan to its manager.

  • starting_type: Name of the manager from which the starting state should be retrieved. Valid values are joint state and pose. If left empty, the starting state will be retrieved from the userdata.

  • starting_name: Name of the starting state. It can be the name of either a pose or a joint state which is located either in a manager or in the userdata. If this slot and starting_type are left empty, the current state of the robot (when executing ComputePlan) will be used as starting state.

Example of use

You can find a concrete example of how to use this state, here.