cambel07's profile - activity

You could start with the MoveIt tutorials. After that, you can check the official universal robots repository that inclu

You could start with the MoveIt tutorials. After that, you can check the official universal robots repository that inclu

SimpleActionClient(c++) lost goals Hi, We are working with a multi-thread motion plan/execution using ActionServers (no

If you can provide access to your files I may have a look. Otherwise, you can check my approach to having dual robots wi

Well it seems you are getting confused with the use of {prefix}, unless you are going to have several robots, you can tr

I think it is exactly as you say, I mentioned {prefix}robot_base_link above because that is the convention that you seem

I think it is exactly as you say, I mention {prefix}robot_base_link above because that is the convention that you seem t

Trajectory swept volumes OCTOMAP / OCTOTREE Hi, I would like to compute the swept volume of a trajectory followed by a

How can I plan ONLY multiple linear trajectories that are independent of each other?

For my use case, I want to plan several trajectories starting from different states and they are unrelated of each other. I am using the SimpleActionClient actionlib.SimpleActionClient("/sequence_move_group", moveit_msgs.msg.MoveGroupSequenceAction). When planning the first trajectory, the result is okay, however when I send the second trajectory to be planned I get the following error:

[ERROR] [1623032769.603690338]: > Planning pipeline threw an exception (error code: -17): Only the first request is allowed to have a start state, but the requests for group: "b_bot" violate the rule

As I understand, the "SequenceManager" seems to interpret the second trajectory as part of the first and then it complains that only the first request is allowed to have a start state. However, how do I tell the "SequenceManager" that the two trajectories are not part of the same sequence and that just treat the second request as a new request?

If I use the SimpleActionClient with the planning_options.plan_only = False, the first trajectory is planned and executed correctly, then the second trajectory is planned and executed correctly. However if the request is planning_options.plan_only = True the I get the above mentioned error. I do not want to plan1->execute1->plan2->execute2, I want to plan1->plan2->execute1->execute2

For both trajectories I was using the move_group to construct the MotionSequenceRequest. However, for the first trajecto

Failed to find root link: Two root links found: [robot_base_link] and [world] With Gazebo, you can not have multiple r

My bad, for both trajectories I was using the move_group to construct the MotionSequenceRequest. However, for the first

My bad, for both trajectories I was using the move_group to construct the MotionSequenceRequest. However, for the first

MoveGroupSequence PILZ plan only multiple unrelated trajectories How can I plan ONLY multiple linear trajectories that a

Hi,

I have a problem when I try to call a Trigger service from Python. There is a service call /ur_hardware_interface/dashboard/program_state of the Universal Robots Driver, I can successfully call the service using the terminal rosservice call /ur_hardware_interface/dashboard/program_state. However when I try to call the service using a simply Python script, it fails

import rospy
from std_srvs.srv import Trigger, TriggerRequest

rospy.wait_for_service('/ur_hardware_interface/dashboard/program_state')
check_connection = rospy.ServiceProxy('/ur_hardware_interface/dashboard/program_state', Trigger)

req = TriggerRequest()
res = check_connection(req)
print("Check connection", res)

I get the following error message:

rospy.service.ServiceException: unable to connect to service: remote error reported: client wants service /ur_hardware_interface/dashboard/program_state to have md5sum 937c9679a518e3a18d831e57125ea522, but it has 522d35a6f2ebd9702cb1e33489e1aa96. Dropping connection.

The universal robot driver is compiled from source in my catkin_ws, it is up-to-date, and has no modifications. The service and the client are running on the same PC so I don't really understand, though just in case I rebuild my catkin_ws but it does not change anything. Additionally, other services published by the Universal Robots Driver do work without problem from the Python script

I see, thank you!

md5sum issue when calling service from Python Hi, I have a problem when I try to call a Trigger service from Python. Th

Hi!

I am trying to use MoveIt's python interface to plan cartesian trajectories with a UR3e robot using ROS-Melodic. I have set up everything and it works fine. I am using it with the TRAC-IK solver.

The problem is that when I get a solution from the planner, the solution can change very drastically from a small motion to an unnecessarily long trajectory. The main difference between trajectories is the goal joint configuration from the goal pose. So my question is, how to request the closes IK solution to the current joint configuration?

I can do that manually by computing the IK solutions using TRAK-IK, computing the closes solution to my current configuration, and then requesting MoveIt to plan using the joint configuration that I choose rather than using a goal pose. When I do that, the resulting trajectory is always similar. However, isn't there a way to request MoveIt to use a specific IK solution over another?

Edit: Video for clarity. From the same initial position, and to the same target pose (Cartesian goal), different IK solutions. https://youtu.be/7GiPZUdlPNA

I see, yes, it would be rather nice to have an option in MoveIt to obtain the IK solution that matches some criteria, su

@gvdhoorn I see, thanks. I suppose for the TRAC-IK solver that is the right solution. should we close this question? sho

MoveIt - Python interface - How to request closest IK solution to current joint configuration? Hi! I am trying to use M

I struggled a lot to do the same. I was not able to compile everything into python3 because of the many specific depende

I struggled a lot to do the same. I was not able to compile everything into python3 because of the many specific depende