The site is read-only. Please transition to use Robotics Stack Exchange
| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
 | 1 | initial version |
Is there a limit on the maximum number of waypoints one can have?
as far as I know the implementation of computeCartesianPath(..) in MoveIt: no, there is no limit. Processing your waypoints will get progressively slower, but that is to be expected.
Does anyone have any idea why it suddenly stops calculating trajectories around point 115
In my experience computeCartesianPath(..) fails when it encounters either: a problem with IK, or if -- at the end -- the computed path violates your jump threshold.
Without more (debug) output it's not necessarily easy to see which of the two it is.
If I may make a suggestion: while MoveIt can do these kind of things, I would suggest you take a look at descartes. It is much better suited for generating tool paths from dense Cartesian trajectories and also supports things like optimisation of trajectories when they are underdefined (ie: 5 dof process with a 6 dof robot).
| | 2 | No.2 Revision |
Is there a limit on the maximum number of waypoints one can have?
as far as I know the implementation of computeCartesianPath(..) in MoveIt: no, there is no limit. Processing your waypoints will get progressively slower, but that is to be expected.
Does anyone have any idea why it suddenly stops calculating trajectories around point 115
In my experience computeCartesianPath(..) fails when it encounters either: a problem with IK, or if -- at the end -- the computed path violates your jump threshold.
Without more (debug) output it's not necessarily easy to see which of the two it is.
If I may make a suggestion: while MoveIt can do these kind of things, I would suggest you take a look at descartes. It is much better suited for generating tool paths from dense Cartesian trajectories and also supports things like optimisation of trajectories when they are underdefined (ie: 5 dof process with a 6 dof robot).
| | 3 | No.3 Revision |
Is there a limit on the maximum number of waypoints one can have?
as far as I know the implementation of computeCartesianPath(..) in MoveIt: no, there is no limit. Processing your waypoints will get progressively slower, but that is to be expected.
Does anyone have any idea why it suddenly stops calculating trajectories around point 115
In my experience computeCartesianPath(..) fails when it encounters either: a problem with IK, or if -- at the end -- the computed path violates your jump threshold.
Without more (debug) output it's not necessarily easy to see which of the two it is.
If I may make a suggestion: while MoveIt can do these kind of things, I would suggest you take a look at descartes. It is much better suited for generating tool paths from dense Cartesian trajectories and also supports things like optimisation of trajectories when they are underdefined (ie: 5 dof process with a 6 dof robot).
Edit:
where it will stop calculating poses but still return a success for some reason
quick comment on this btw: computeCartesianPath(..) will almost always 'succeed', but the return value (fraction) should be used to determine whether it completely processed your list of waypoints. Anything < 1.0 basically means that it couldn't successfully generate a path through all the poses in the list.
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
