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 |
What I did to get this working again with Kinetic was to create a new node that wraps the actionlib processing, so in my voice head test node I publish a "speech" topic to this new node that is the string of text to speak. The new node gets the string from the topic and makes the calls to actionlib to speak. Then when the actionlib result is available I parse the result and publish a message for my test application telling how long the operation took. Likewise I publish a wav topic passing the path to a wave file to play.
This seems to work quite well, So if someone else should hit the same issue that I was having with actionlib and Kinetic this should resolve it.
What was I using actionlib? I needed to get an approximation of how long the speech took. Since I need to block voice recognition while the robot is speaking, this works well. And eliminates the robot trying to respond to any speech that it has output.
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.
