ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
2020-04-16 08:10:52 -0500 | received badge | ● Famous Question (source) |
2019-05-09 09:00:19 -0500 | received badge | ● Taxonomist |
2015-04-30 22:47:51 -0500 | received badge | ● Notable Question (source) |
2015-03-16 03:57:18 -0500 | received badge | ● Notable Question (source) |
2015-02-16 12:27:39 -0500 | received badge | ● Popular Question (source) |
2015-02-03 15:01:03 -0500 | commented answer | Nodes of robotic system and actions Thank you, you made it very clear. |
2015-02-03 15:00:34 -0500 | received badge | ● Scholar (source) |
2015-02-03 10:49:09 -0500 | received badge | ● Popular Question (source) |
2015-02-03 10:23:47 -0500 | commented answer | Nodes of robotic system and actions Thanks for your nice and clear response. So, I need to implement a communication channel between the clients in order they know what each other do (ie the hand client to know what feedback from the arm has the arm client). Perhaps, a topic? |
2015-02-03 07:01:24 -0500 | asked a question | Nodes of robotic system and actions Maybe I do not fully understand how actionlib works, but I have a pretty basic question. Suppose you have a robotic system, with an arm, a hand attached to the arm, a video camera etc, that we want to work together. Scenario: The camera see the object, the arm moves towards the objects and the hand is touching the object (we are getting sensor feedback from the hand). Finally, the hand grasps the object. What is the model for this using actions? 3 clients-1server? or 1 client-3servers for each component? In the schema 3 clients-1server, can we control the arm and the hand simultaneously or we have to cancel a goal to the arm, to send another for the hand? As far as I know an action_server represents a node, so I suppose we need 3 servers. But how can we coordinate the actions between the servers? |
2014-12-17 08:13:37 -0500 | received badge | ● Supporter (source) |
2014-12-17 07:19:32 -0500 | asked a question | Actionlib and grasping Hi everyone, I have built a library in C++ with which I control a robotic hand. So I have some classes with some respective functions. For instance there is a function closeHand(), or moveFinger(position), nothing fancy. I would like to to run this code under ROS and I suppose I have to use actions. I tried the tutorials for actionlib regarding fibonacci sequence, but I couldn't totally comprehend how to generalize it and using it for my project. I am really new to this so, maybe, thats the case. For instance, I cannot understand if I have to use more than one servers (one for each of my functions), or what type of goals can I have. If I use one server, can I have multiple goals in the action file? |