ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
2021-10-25 12:01:04 -0500 | received badge | ● Popular Question (source) |
2021-10-25 02:51:12 -0500 | commented answer | catkin: When are the install directives in CMakeLists.txt effective? Thanks for answering. Although this is explaining how to install the files (and where), it doesn't mention why the insta |
2021-10-25 02:51:12 -0500 | received badge | ● Commentator |
2021-10-25 02:51:05 -0500 | commented question | catkin: When are the install directives in CMakeLists.txt effective? @gvdhoorn Partially. I'm also asking the necessity of installation, if the package works without doing that. |
2021-10-25 00:08:38 -0500 | asked a question | catkin: When are the install directives in CMakeLists.txt effective? catkin: When are the install directives in CMakeLists.txt effective? I tested a simple catkin package, and have some que |
2021-10-24 22:57:07 -0500 | commented question | rosmake error [melodic - ubuntu 18.04] @gvdhoorn 1. rosmake cannot build catkin packages, of course. But packages build with rosmake can work with catkin-built |
2021-10-22 01:14:56 -0500 | commented question | rosmake error [melodic - ubuntu 18.04] @jarvisschultz Thanks for the comments. Yes, deprecation is too strong, if there is no agreements. I understand the adva |
2021-10-20 23:27:46 -0500 | commented question | rosmake error [melodic - ubuntu 18.04] @jarvisschultz Where is the evidence that rosmake was deprecated? In its official ROS Wiki, the status is maintained, an |
2021-10-13 13:22:00 -0500 | received badge | ● Famous Question (source) |
2021-04-07 17:49:44 -0500 | received badge | ● Good Question (source) |
2021-03-31 10:17:26 -0500 | received badge | ● Notable Question (source) |
2020-11-08 08:48:21 -0500 | received badge | ● Popular Question (source) |
2020-11-01 21:46:45 -0500 | asked a question | Elapsed time goes weird Elapsed time goes weird Hello, I'm experiencing periodic Moved backwards in time errors from robot_state_publisher. I'm |
2019-07-06 11:43:00 -0500 | received badge | ● Nice Question (source) |
2019-03-27 02:31:17 -0500 | received badge | ● Famous Question (source) |
2018-09-14 17:28:04 -0500 | marked best answer | Motoman FS100: Velocity feedback control Hello, I'm controlling Yaskawa Motoman SIA10F (controller: FS100) over ROS. I'd like to do a velocity feedback control (PD control). As the first-step implementation, I made a simple joy stick control: joy stick command is converted to target joint velocities. My code is like: The current joint velocities are computed internally, because FS100 does not publish joint velocities in the When executing this code, it did not work, and I encountered errors: How can I implement a velocity feedback control? If there is an example code, please let me know. Thanks! |
2018-04-18 22:47:44 -0500 | received badge | ● Famous Question (source) |
2018-02-26 04:07:11 -0500 | received badge | ● Notable Question (source) |
2018-01-11 20:07:40 -0500 | marked best answer | Collision detection in Python Hello everyone! I would like to ask a simple question. How can we detect collisions in Python? (as simple as possible) My goal is as follows:
I found several solutions as follows, but I want simpler solutions.
Hence, to the best of my knowledge, I need to implement a python interface to call C++ functions (by using either the Python/C interface or the ROS communication), which sounds a bit complicated. If anyone can suggest simpler solutions, please tell me. Thanks in advance. |
2017-12-04 01:24:15 -0500 | received badge | ● Popular Question (source) |
2017-12-03 00:49:22 -0500 | received badge | ● Supporter (source) |
2017-12-01 05:14:03 -0500 | asked a question | Motoman FS100: Velocity feedback control Motoman FS100: Velocity feedback control Hello, I'm controlling Yaskawa Motoman SIA10F (controller: FS100) over ROS. I' |
2017-11-28 16:30:44 -0500 | received badge | ● Notable Question (source) |
2017-11-27 18:21:11 -0500 | marked best answer | Motoman FS100: Missing velocity and effort in joint_states I'm controlling Yaskawa Motoman SIA10F (controller: FS100) over ROS. Previously the What I am doing to launch the Motoman system is as follows: |
2017-11-27 18:21:11 -0500 | received badge | ● Scholar (source) |
2017-11-27 18:17:49 -0500 | commented answer | Motoman FS100: Missing velocity and effort in joint_states Mmm it could be my mistake. What I'm sure is that I saw a "response" of robot for pushing it. I thought the response w |
2017-11-27 08:19:44 -0500 | received badge | ● Popular Question (source) |
2017-11-27 00:16:31 -0500 | commented question | Motoman FS100: Missing velocity and effort in joint_states I added some packages, but now I removed them. |
2017-11-26 23:46:38 -0500 | asked a question | Motoman FS100: Missing velocity and effort in joint_states Motoman FS100: Missing velocity and effort in joint_states I'm controlling Yaskawa Motoman SIA10F (controller: FS100) o |
2017-10-10 04:48:51 -0500 | received badge | ● Famous Question (source) |
2017-04-04 06:41:59 -0500 | received badge | ● Nice Question (source) |
2017-01-18 22:04:32 -0500 | received badge | ● Popular Question (source) |
2017-01-18 22:04:32 -0500 | received badge | ● Notable Question (source) |
2016-05-08 15:36:43 -0500 | marked best answer | Collision is not detected with cylinder model I am testing the arm_navigation's collision detection for a virtual object added to the planning scene and a virtual object attached to the robot. When using CYLINDER models, the collision detection fails; the results are OK when using BOX models. Using CYLINDER models is better in my case, so if somebody knows the reason, please tell me. Here is the detail:
First, I added a virtual CYLINDER object to the planning scene through the /environment_server/set_planning_scene_diff service. Also, I attached a virtual CYLINDER object to the robot's left gripper. Then, I checked the collision during moving the robot's arm through the planning_scene_validity_server/get_state_validity service. It could detect collisions between 1. the robot's body and the virtual scene object, and 2. the virtual robot's object and the robot's body. However, it could not detect collisions between the virtual robot's object and the virtual scene object. I tested to change the SetPlanningSceneDiffRequest.operations.collision_operations parameter, but it did not work. Here is a video to show this problem: http://youtu.be/oRmfEwKrD9s (CYLINDER case) By changing the objects to BOX models, it works as I expected. The video is here: http://youtu.be/YH-MAvbKm9M (BOX case) Many thanks! Update: Combinations of (CYLINDER and BOX) or (BOX and CYLINDER) work well. Only the (CYLINDER and CYLINDER) case is weird. |
2016-04-22 16:10:07 -0500 | asked a question | MoveIt! Collision Detection of Single Point Hello, I am using planning_scene::PlanningScene of MoveIt! to detect collisions with the checkCollision member function. Now I would like to program a lower-level collision check. How can I detect a collision of a single point (x,y,z)? My goal is for a given set of points (i.e. point cloud), categorizing each point into (1) colliding with a robot or an object or (2) no-collision. I guess a collision detection of single point would solve this problem. But if have any other ideas, please let me know. Thanks! |
2015-08-21 07:57:31 -0500 | received badge | ● Famous Question (source) |
2015-07-21 18:56:22 -0500 | commented question | Error: Links 'torso_lift_link' and 'l_wrist_roll_link' do not form a chain I haven't solved this problem yet, but my tentative solution would work in many cases. Use an external PC and connect to the same network as PR2. Then launch a node using above code (e.g. a planner, state-validity checker) on the external PC. Now you can use that node as if it is working on PR2. |
2015-05-12 10:36:10 -0500 | received badge | ● Famous Question (source) |
2015-05-08 02:41:55 -0500 | received badge | ● Notable Question (source) |
2015-05-07 18:30:21 -0500 | received badge | ● Popular Question (source) |
2015-03-10 20:45:05 -0500 | received badge | ● Self-Learner (source) |
2015-03-10 20:45:05 -0500 | received badge | ● Necromancer (source) |
2015-03-10 20:45:05 -0500 | received badge | ● Teacher (source) |
2015-03-10 19:13:36 -0500 | answered a question | Collision is not detected with cylinder model The collision detection of planning_environment ( http://wiki.ros.org/planning_environment ) seems to be using the collision checker of ODE (open dynamics engine). The codes are not available online now, but they are:
Anyway, unless fixing this bug, collision detection between two cylinder models does not work. Hence, a better solution is using other collision detection library, such as:
Though MoveIt! is an integrated planning tool, we can use it as a collision detection library (and it is easy to use on ROS). The following videos are the similar situations where we are using MoveIt! as a collision checker. No problem is happening in the collision detection between two cylinder models case.
|
2015-03-10 17:04:31 -0500 | asked a question | Error: Links 'torso_lift_link' and 'l_wrist_roll_link' do not form a chain Hello everyone. I have a problem in using MoveIt! and the real PR2 robot. I am using MoveIt! for collision checking from C++. When I load a robot model with robot_model_loader::RobotModelLoader from the "robot_description" parameter on the ROS parameter server with a code like: the program outputs an error and dies as follows: I tested the same code on the gazebo simulator where the program could run correctly, then I executed the code on the real PR2 where the above error happened. I checked that the "robot_description" parameter provides a correct information as follows: Hence, the problem will be a library or a configuration file. However, I could not identify the problem. I checked the URDF and SRDF of PR2, but they are the same as those of the external PC. Could you please give me any advises to solve this problem? Thank you so much in advance, Edited@2015-05-07 I still can not solve this problem. I also tested to execute a tutorial program such as collision_contact_tutorial.launch in pr2_moveit_tutorials: but there was the same error (Links 'torso_lift_link' and 'l_wrist_roll_link' do not form a chain). By the way, those links are used in the beginning of the SRDF file as: Anyway now I doubt this will be a moveit!'s problem. Has someone experienced a similar problem? |
2015-02-25 13:01:45 -0500 | commented question | Collision detection in Python Good idea. I have created a topic about this. |
2015-02-25 01:59:16 -0500 | received badge | ● Notable Question (source) |
2015-02-24 12:59:29 -0500 | received badge | ● Popular Question (source) |