Open Software. Open Knowledge.


Tim's WebLog

Fedora at RoboCup

Fedora RobotWe arrived on Sunday in Suzhou, China. All went pretty smoothly and there are dozens of volunteers and staff that makes it a pleasant stay. There is a 20 min bus shuttle that we take every morning to the venue. A continuously updated short report is on our RoboCup website.

As you can see on the photo we have put Fedora stickers at prominent places on our robots (there is one at the back side as well). It fostered some interest. I was talking to a member of the executive committee for the RoboCup@Home league. He liked the idea of a LiveCD with robotics software on it. A few other guys were asking "what is that T-Shirt about?"... Quite cool, hope to spawn more interest in Fedora in the RoboCup community.

The first tests didn't went as smooth as expected. In the first test, where the robot introduces itself to the audience and other teams, showing its abilities, sensors and actuator we needed a restart. A textile cover of the robot hang into the laser scanner which put it into "escape mode", a mode where it's moving forward slowly to get away from the obstacle. After the restart it went smooth but time was short. The fast follow went pretty well. Masrur "ran" through the environment with the robot following. Unfortunately the other team from Koblenz didn't make it in time so they started late an we couldn't demonstrate our passing capabilities. So after the first two tests we were first.

The next two tests we screwed up because of wifi problems. Yesterday we had the lost and found test, where the robot has to find three objects that are randomly placed in the environment. We implemented a new object detection and were quite keen to see it in action. But then the wifi broke down for no visible reason and we couldn't start the application... Today we did the who-is-who challenge. Also a challenge were we introduced new code, now for face detection and recognition and using a new way to formulate the application. We had wifi problems - again. Just after that we found out that it was the microphone that was put on the robot to make the text-to-speech output heard. We did a test with a busy ssh session. If you turned the microphone on the session would immediately stall, and when you turned it off it would continue immediately. The other microphone isn't as bad, this is why it worked in the first two tests flawlessly...

After the first days with two screwed up tests we are currently third (results). Tomorrow will be the open challenge, where every team can demonstrate what it can do best with their robot. After that the stage 2 tests will start where only 10 of the attending 14 teams are allowed to compete. From these 5 teams will compete in the finals on Sunday. Let's see how this works out.

Top 5 Pages
Friends' blogs
SquidGuard Webmin Module

Palm Software
UniMatrix UniMensa UniSorter
UniChat OHS Mobile Onager

My Bookshelf

Valid XHTML 1.1!

RSS Copyright © 2000-2016 by Tim Niemueller