Javier J. Salmerón García's profile - activity

Hello everyone,

I am using gazebo 1.9 with ROS Groovy. I want to add an erratic robot in the willowgarage world. In order to do so I have the follwing launch file:

<launch>

  <!-- We resume the logic in empty_world.launch, changing only the name of the world to be launched -->
  <include file="$(find gazebo_ros)/launch/empty_world.launch">
    <arg name="world_name" value="worlds/willowgarage.world"/> <!-- Note: the world_name is with respect to GAZEBO_RESOURCE_PATH environmental variable -->
    <arg name="paused" value="false"/>
    <arg name="use_sim_time" value="true"/>
    <arg name="gui" value="true"/>
    <arg name="headless" value="false"/>
    <arg name="debug" value="false"/>
  </include>

<param name="robot_description" command="$(find xacro)/xacro.py $(find erratic_description)/urdf/erratic.urdf.xacro" />

<node name="spawn_urdf" pkg="gazebo_ros" type="spawn_model" args="-param robot_description -urdf -model erratic" />

</launch>

This example file worked perfectly with pr2, but when I launch it I get the following error:

[rospack] Error: stack/package smart_arm_description not found
[librospack]: error while executing command
[FATAL] [1374747130.949047781]: Package[smart_arm_description] does not have a path
[spawn_urdf-4] process has finished cleanly

I cannot find this smart_arm_description package, can someone help me?

Thank you in advance

Hello everyone,

I am developing an application with several nodes: one receives a message (for example: ImageConstPtr&) that will be forwarded to another node in particular (via another topic), following a pool of workers approach. Can I create a publisher that use its function publish() with that ConstPtr& message, or should I transfer its contents to a new message object?

Thank you in advance

Hello everyone,

I am developing a parallel solution with ROS, and I want to develop a node that executes certain functions when a another node in the architecture is added or destroyed at runtime. I know that I can do some status polling with rosgraph library, but is there any way to have a callback function when any of the other nodes execute ros::init or ros::shutdown?

Thank you in advance

Hello everyone,

I want to install libavcodec-extra-53 instead of libavcodec53 (AFAIK the latter should include everything from the former plus some other codecs). However, when I try to do so it forces me to remove all my ROS installation. Is there any way to make ROS compatible with libavcodec-extra-53?

Thank you in advance

Hello everyone,

Following the tutorial in Ros Wiki for creating a simple image publisher (from a set of .png files) http://ros.org/wiki/image_transport/Tutorials/PublishingImages I found a compilation error in groovy (in fuerte everything works fine):

cv::WImageBuffer3_b image(cvLoadImage(filePath.c_str(), CV_LOAD_IMAGE_COLOR));
sensor_msgs::ImagePtr msg = sensor_msgs::CvBridge::cvToImgMsg(image.Ipl(), "bgr8");

This is the error I found:

 error: ‘sensor_msgs::CvBridge’ has not been declared

I suppose that the API has changed, but I cannot find an alternative way. Can somebody help me?

Thank you in advance

Hello everyone,

I am executing a multi-node ROS (ROS Fuerte Ubuntu 12.04) architecture. These nodes are spread in different machines. When a certain node finishes, I want the rest to be killed. I thought that using required=true parameter in the roslaunch would be sufficient. However, I found that, when the node finishes, only the nodes located in the same machine would be killed as well. Is there any way to make all the nodes in all the machines be killed (I mean automatically, not using Ctrl-C in the roslaunch).

Thank you in advance,

Hello everyone,

I have a newbie problem here. I am trying to implement a ROS Architecture which is similar to a producer-consumer problem. I have one node which is publishing images and a group of nodes that would take the image and do some processing in it. The behaviour that I want to implement is the following: If a consumer node takes one image, I do not want the others to receive that same image. Using the default image_transport publisher, I suppose that all the subscribers will receive the same image at a time. How can I avoid this?

Thank you in advance

I am doing an experiment that consists on transferring JPG compressed stereo pairs from one node to another at 5Hz. I am measuring the message transfer time using two technologies: UDPROS and TCPROS. The frame sizes tested are 320x240 and 640x480.

The point is that the average message transfer times are better when using TCPROS. In the best case UDPROS matches the transfer times obtained with TCPROS (Very rarely UDPROS gets a little better). How is that possible? Am I doing something wrong?

The technology used is 802.11ac WiFi.

I tried with 8x8 frames and it works. Therefore you must be right, 320x240 raw frames are to big to be sent with UDPROS.

I am developing a robotic application that sends raw stereo images (computer A) using the image_transport library. Another application receives them (computer B) and synchronizes the stereo pairs and camera infos. This application has always worked perfectly using the default transport hints.

Today I changed the transport layer from TCPROS to UDPROS, and now the receiver node only gets the camera infos, but not the frames. I tried changing the datagram size, the connection technology (Wifi, Gigabit...) and the computers, without any success. The smallest frame size I tested was 320x240 and the send rate was also changed to 1 Hz.

When both nodes are executed in the same computer (e.g. computer A), everything works perfectly, so I suspect that it is not a programming issue.

Any ideas? Thank you in advance