MoveIt is currently primarily a kinematic motion planning framework - it plans for joint or end effector positions but not velocity or acceleration. However, MoveIt does utilize post-processing to time parameterize kinematic trajectories for velocity and acceleration values. Below we explain the settings and components involved in this part of MoveIt.
By default MoveIt sets the velocity and acceleration of a joint trajectory to the default allowed in the robot’s URDF or
joint_limits.yaml is generated from the Setup Assistant and is initially an exact copy of the values within the URDF. The user can then modify those values to be less than the original URDF values if special constraints are needed. Specific joint properties can be changed with the keys
max_position, min_position, max_velocity, max_acceleration. Joint limits can be turned on or off with the keys
The speed of a parameterized kinematic trajectory can also be modified during runtime as a fraction of the max velocity and acceleration set in the configuration values, as a value between 0-1. To change the speed on a per-motion plan basis, you can set the two scaling factors as described in MotionPlanRequest.msg. Spinboxes for setting both of those factors are also available in the MoveIt MotionPlanning RViz plugin.
Time Parameterization Algorithms¶
MoveIt can support different algorithms for post-processing a kinematic trajectory to add timestamps and velocity/acceleration values. Currently there are three available by default in MoveIt:
- Iterative Parabolic Time Parameterization
- Iterative Spline Parameterization
- Time-optimal Trajectory Generation
The Iterative Parabolic Time Parameterization algorithm is used by default in the Motion Planning Pipeline as a Planning Request Adapter as documented in this tutorial. Although the Iterative Parabolic Time Parameterization algorithm MoveIt uses has been used by hundreds of robots over the years, there is known bug with it.
The Iterative Spline Parameterization algorithm was merged with PR 382 as an approach to deal with these issues. While preliminary experiments are very promising, we are waiting for more feedback from the community before replacing the Iterative Parabolic Time Parameterization algorithm completely.
Time-optimal Trajectory Generation introduced in PRs #809 and #1365 produces trajectories with very smooth and continuous velocity profiles. The method is based on fitting path segments to the original trajectory and then sampling new waypoints from the optimized path. This is different from strict time parameterization methods as resulting waypoints may divert from the original trajectory within a certain tolerance. As a consequence, additional collision checks might be required when using this method.
Open Source Feedback
See something that needs improvement? Please open a pull request on this GitHub page